Who pays for the roof — mandates, subsidies, and repeal

When greenery goes onto a roof, the first thing that moves is money. The waterproofing is redone, a root barrier is laid, soil is hauled up, irrigation pipe is run. All of it is fronted by the building's owner. Yet when you itemise what comes out of it, only a small share stays in the owner's hands. According to the US Environmental Protection Agency, a green roof can hold its surface as much as 56 degrees Fahrenheit — around 31 degrees Celsius — below a conventional roof, lower nearby air temperatures by up to 20 degrees Fahrenheit, and cut cooling loads by as much as 70 percent. Of these, the saved cooling bill and the extended life of the membrane belong to the owner; the cooler air outside and the rain that never reaches the sewer never enter the owner's account.

Economists call this an externality: the party bearing the cost and the party receiving the benefit are not aligned. Leave the misalignment in place and a rational owner declines to green. Desirable for the city, not worth it for the individual building — that single fact is why rooftop greening has almost never been discussed apart from regulation. Behind the claim in our first instalment that regulation rather than the market made the rooftop green lies exactly this accounting structure. In Germany, some 90 percent of newly created green roofs are said to arise from ecological compensation requirements attached to development. They are not what a market chose of its own accord.

This article therefore reads regulation not as principle but as sums. What burden did the mandate impose? What share of it did the subsidy fill? How much does a fee reduction return in a year? And what actually feeds the people who sell crops from a roof as a business? With the figures laid out, we return to what happened in Toronto in October 2025. When the repeal of a rule is explained as cutting unnecessary costs, can we tell precisely which costs were counted and which benefits were not?

There are broadly three ways to close an externality, on a roof or anywhere else. The first is the mandate: require greening on buildings meeting certain conditions and leave the cost with the owner. This is what Toronto adopted in 2009 and Tokyo in 2001. The second is the subsidy: have the public purse carry part of the construction cost and lower the owner's burden. The third is the price: put a value on the benefit and return it to the owner, for instance by reducing the sewage charge on a greened building. The three are not exclusive, and most real cities combine at least two.

What matters is that the three work differently and fail differently. The mandate is the most powerful and stacks up area fastest, but it accumulates resentment among those made to pay and can be repealed politically. The subsidy provokes less resistance but spreads only as far as the budget allows and bends to a single year's fiscal weather. The price is the most durable, but its sums are usually far too small to repay the construction cost. In what follows we take all three down to their actual figures and check how far each really reached.

Toronto's green roof bylaw was passed in May 2009, the first in North America to require green roofs on new buildings. It applied to new commercial, institutional and residential construction with a gross floor area of 2,000 square metres or more, and required greening of between 20 and 60 percent of the available roof space on a sliding scale by building size. Available roof space was defined not as the whole roof but as the area left after subtracting space given to renewable energy installations and to private terraces and residential outdoor amenity area, up to a maximum of two square metres per unit. The weight of the duty, in other words, was tuned to what the roof was otherwise for. By the city's own count, more than 1,200 roofs were greened under the bylaw.

Tokyo's duty takes a different shape. The Ordinance on the Protection and Restoration of Nature in Tokyo came into force in April 2001, and its Article 14 requires anyone constructing or extending a building on a site of 1,000 square metres or more within the metropolis — 250 square metres or more where the land is held by the national or a local government — to file a greening plan. For roofs, greening of roughly 20 percent of the usable portion is required, rising to 30 percent where schemes such as the comprehensive design system apply. Toronto pulled its trigger from a building's floor area; Tokyo pulls it from the site area. The difference looks minor and is not: it sorts an entirely different set of buildings into the net.

And there is one further, decisive difference. Toronto's bylaw provided an escape: pay money instead of greening. Tokyo's system provides none. Either you green as the filed plan says, or you do not; there is no route by which the duty can be converted into cash. As the next section shows, that single design choice separated the two cities sharply in how durable and how self-financing their regimes turned out to be. A mandate is not only a question of what you require, but of whether you let the requirement be paid off.

Toronto's bylaw allowed a developer, with approval, to pay 200 Canadian dollars per square metre instead of building the required green roof — the familiar cash-in-lieu arrangement. At first glance this hollows the duty out; look at where the money went and the judgement changes. The payments funded the city's Eco-Roof Incentive Program, which paid out grants of 100 Canadian dollars per square metre, to a maximum of 100,000 dollars, for green roof work. Cool roofs drew 2 to 5 dollars per square metre up to 50,000, and up to 1,000 dollars was available for the structural assessment that tells an existing building whether it can carry a green roof at all.

The grants went to buildings the bylaw did not reach: existing structures, new construction below 2,000 square metres of gross floor area, and new projects by the city's school boards and by not-for-profits. Taken as a whole, then, the regime drew money from the large new developments that carried the duty and redistributed it to the small buildings, the schools and the non-profits that did not. The mandate financed the subsidy, and the subsidy covered the ground the mandate could not reach. As rooftop policy goes, that was a remarkably well-closed loop.

The payment has another face, however. A price of 200 Canadian dollars per square metre also works as a ceiling on the cost of escaping. For any building whose green roof would cost more than 200 dollars a square metre, paying is cheaper — which means the figure set the effective upper bound of the duty. The numbers cited by the English Wikipedia put a properly designed and installed extensive green roof at 108 to 248 dollars per square metre and an intensive one at 355 to 2,368. Setting aside differences of currency and year, the rate of 200 sat squarely in the middle of the extensive band. The regime imposed a duty and, in the same breath, priced it.

To think about subsidies you first need the size of the hole they are filling. The German building-greening federation BuGG gives a guide figure of 30 to 60 euros per square metre for an extensive green roof and from 80 euros for an intensive one. That is close to an order of magnitude below the North American numbers, because materials and installation have matured into an industry and the extensive system is standardised. The FBB reports that some eight million square metres of roof are newly greened each year in Germany, more than 80 percent of it extensive. Some German cities subsidise up to half of that construction cost.

Japan's grants are far more generous per unit and far smaller in scale. Mitaka's greening grant, for instance, covers half the actual cost of a rooftop planting, up to 20,000 yen per square metre; wall greening draws up to 10,000 yen per square metre, and roof and wall together are capped at one million yen. To qualify, the greened area must be at least one square metre, a fall-prevention barrier of at least 1.1 metres must be installed on the roof, and the building must be strong enough to carry the new load and hold a certificate of inspection under the Building Standards Act. The unit rate is high, but a ceiling of one million yen buys only about fifty square metres.

Here the shape of Japan's arrangement comes into view. The duty is imposed by the metropolitan government; the money comes from the wards and cities. And because the metropolitan ordinance has no payment-in-lieu mechanism, there is no route by which the mandate could finance the grants as it did in Toronto. Two regimes, run by different authorities out of different purses for different purposes. As the first instalment showed, Japan's greened roof area rose with the ordinances of the early 2000s. What raised it was the mandate; what has carried it since is annual municipal budgeting. The combination succeeded in producing a step change in area. It was never designed to keep producing one.

The third instrument prices the benefit and returns it to the owner. Many German municipalities levy the sewage charge in split form — the gesplittete Abwassergebühr — dividing it into a foul-water component and a rainwater component, the latter assessed on the impermeable area of the site. Because a green roof holds rain and reduces runoff, it is treated favourably in that assessment and the charge falls. Measure the benefit, translate it into a price, return it to the party responsible: in theory, the textbook solution to an externality. So how much comes back? The figures collected by the German Wikipedia give an average of about 0.46 euros per square metre a year, and 1.12 euros at most, in Cologne.

Set that against the cost of the work. If an extensive roof runs 30 to 60 euros per square metre, then at the average reduction of 0.46 euros it takes, on a simple calculation, something like 65 to 130 years to recover. Even at Cologne's maximum of 1.12 euros it is 27 to 54 years. Measured against the life of a waterproof membrane, that is fairly called unrecoverable. The calculation looks only at the fee, of course, and excludes benefits the owner keeps for himself, such as the saved cooling bill and the longer-lived membrane. Even so, the fact is heavy: when the most cleanly measurable benefit of all, rainwater, is given a price, what comes back is a few percent of what the work cost.

The real work of the price instrument is probably not repayment but signalling. The institutional declaration that a greened roof will not be counted as impermeable tells designers and clients that this is work of public value. More important than the sum itself was that it created an accounting in which rainwater is a liability on the city's books. That some 90 percent of Germany's newly created green roofs are said to arise from ecological compensation requirements attached to development is a consequence of that accounting taking hold. But the same 90 percent says something else too: the signal alone never greened a roof.

So far the question has been who bears the cost of the work. What, then, keeps afloat the business of growing and selling crops on the resulting roof? Brooklyn Grange began in April 2010 and now grows across three rooftops in New York. Annual production runs to roughly 80,000 pounds, about 36 tonnes, of organic vegetables and around 1,500 pounds, some 680 kilograms, of honey. But the company does not sell only vegetables and honey. It also runs a business providing green-roof design, installation and consulting to clients worldwide. It is a company that farms on roofs and, at the same time, a company that makes roofs farmable.

Lufa Farms in Montreal solved it differently. Founded in 2009, it began harvesting in 2011 from a commercial rooftop greenhouse of about 31,000 square feet in the city, serving roughly 200 customers at the start. It expanded to Laval in 2013 (about 43,000 square feet), Anjou in 2017 and Ville Saint-Laurent in 2020 (about 164,000 square feet), and now runs four greenhouses totalling some 300,000 square feet around the city. What distinguishes it is the selling. The company runs an online marketplace that pairs its own produce with goods from local producers, delivering around 20,000 baskets a week to more than 500 pick-up points across Quebec. It employs about 200 people.

What the two share is that the crop itself is not where the profit comes from. One sells the technique of making a roof; the other sells the distribution that links roof to table. What the rooftop gives the business is not yield but story and proximity. And they share one more thing that should not be missed: both use a roof that was already there. Neither bought land. The largest single factor underwriting the economics of a rooftop farm is that the costliest line item in any city — ground rent — is paid thinly as rent, or hardly paid at all. That is a strength and, in the same movement, a fragility: the business can end at the convenience of whoever owns the roof.

In 2025 the government of Ontario repealed Toronto's green roof bylaw. According to reports, the province signalled its intention when it tabled the Fighting Delays, Building Faster Act, 2025 (Bill 60), and the repeal was ultimately effected by an order-in-council signed on 23 October, using a provision of the City of Toronto Act. It is reported to have taken effect in early November. The bylaw was not repealed by the council that wrote it; the enabling provision in the superior statute was removed by the province. Greening reverted to voluntary.

What deserves attention is how the reason was framed. As reported, the office of the province's housing minister explained the move as part of a policy of building infrastructure and homes faster, and said that by making green roofs voluntary, builders and taxpayers gain flexibility of choice while unnecessary costs are cut. What is counted here is the construction cost of a green roof — a figure that fits on one line of a quotation. What is not counted is the value of rain that never reaches the sewer, of air that runs a little cooler, of a membrane that lasts longer. The first appears as a sum paid by somebody identifiable at an identifiable moment; the second spreads thinly across a whole city and sits on nobody's books.

That asymmetry maps exactly onto the figures assembled here. Construction costs — 30 euros a square metre, 248 Canadian dollars a square metre, whatever the currency and level — always exist as a definite number. Meanwhile the one serious attempt to price the benefit, Germany's sewage charge, returned 0.46 euros a year. In a world where benefits are not properly translated into prices, only the cost side comes in a shape that survives a political argument. Toronto's sixteen years and its 1,200-plus roofs could not withstand that asymmetry. The point made in our first instalment, that what regulation made regulation can unmake, is more precisely put like this: a benefit with no price attached is not counted when the cutting starts.

This article has reread rooftop greening not as principle but as the allocation of cost. Its starting point was that the party who pays and the party who benefits are not the same. The owner fronts the whole of the construction cost, while of the benefits the US EPA lists — up to 56 degrees Fahrenheit off the surface temperature, up to 20 degrees off nearby air, up to 70 percent off the cooling load, and runoff cut by about 60 percent on an extensive roof and up to 100 percent on an intensive one — only the cooling bill and the longer-lived membrane land in the owner's account. Against that we compared the three instruments available to regulation, mandate, subsidy and price, each taken down to its actual figures, and then returned to Toronto in October 2025. From definitions to sums, and from sums to repeal: that is the route taken here.

Three points deserve restating by name. First, Toronto's cash-in-lieu was the subsidy's source of funds. The 200 Canadian dollars per square metre was collected from the very buildings the duty covered — new construction of 2,000 square metres or more — and paid out to the ones it did not: existing buildings, new construction below 2,000 square metres, and projects by school boards and not-for-profits, at 100 dollars per square metre up to 100,000 through the Eco-Roof Incentive Program. Second, prices do not repay. Germany's split sewage charge returns an average of 0.46 euros a square metre a year, and 1.12 in Cologne at most, against 30 to 60 euros of construction cost for an extensive roof — a payback of many decades. Third, in Japan the duty and the money sit with different bodies: Tokyo imposes greening on sites of 1,000 square metres or more under Article 14 of its ordinance in force since April 2001, while the money is put up separately by municipalities, as in Mitaka's half-the-cost grant of up to 20,000 yen per square metre and one million yen in total.

There are also things not known, and reservations to state. On Toronto's repeal, the primary documents of the city and the province could not be reached from the environment in which this was written, and the account rests on news reports. That the order-in-council was signed on 23 October 2025, that it took effect in early November, and that more than 1,200 roofs were greened under the bylaw are all reported figures that have not been traced to source. The cost levels need care too: Germany's 30 to 60 euros and North America's 108 to 248 dollars differ in currency, in year and in what they include, and do not bear direct comparison. The payback calculation on fee reductions deliberately excludes saved cooling costs, membrane life and any change in asset value; it is not a verdict on the economics of greening as such. An objection is fair, too: if asset value and tenant preference push towards greening even without a mandate, greening may continue after the repeal. Toronto's grant programme and design standards are in fact reported to remain in place. Whether the repeal means the end of greening is a question the next few years of construction figures should answer, not this article.

As the sixth instalment of the series The roof as farmland, this article has substantiated at the level of money the proposition set out on day one: that regulation, not the market, made the rooftop green. The mandate made the area, the payment in lieu made the subsidy, and the price made the accounting. And of the three, the most powerful vanished the fastest. The final instalment turns to the questions this series has kept in view without yet facing squarely. Does rooftop farming genuinely contribute to a city's food? Have mandates become an alibi for the greenery lost at ground level? And granting every one of those limits, what should a citizen actually do first? The talk of institutions and economics ends here; tomorrow we move to limits and to practice.