Solar spreadsheets

Environmentalism needs a full accounting

"Think global, act local" has been environmentalism's working instruction for half a century. But it has had a structural flaw built into it from the start: a decision made locally is by definition judged locally — against whether this project, on its own, is worth doing. And for almost any well-intentioned environmental accommodation, taken on its own, it is. A bat-friendly railway crossing is better than no bat-friendly crossing. A solar farm on farmland is better, for the climate, than no solar farm at all. Taken one at a time, there is almost nothing local environmentalism can't justify saying yes to.

The maxim needs a third clause: think global, act local, reckon regional. Somewhere between the individual decision and the global outcome, some body needs to keep the regional picture in proportion — not so cautious that nothing gets built, not so generous that the fields are gone before anyone thinks to count them.

A defensible decision, an indefensible result

Start with the result.

A kilometre of railway track in Buckinghamshire runs beneath a curved mesh structure, ten metres high at its tallest, built so that one of Britain's rarest bats can cross safely overhead. The Sheephouse Wood Bat Mitigation Structure — the now-infamous "bat tunnel" — cost around £100 million (in 2019 prices).

Nobody involved seems entirely happy with how this turned out. Natural England, the body responsible for protecting the Bechstein's bat, says it never required this particular structure. The UK's environment secretary called the spending "ludicrous." The Chancellor singled it out as a symbol of everything that makes British infrastructure too slow and too expensive to build. And this wasn't an isolated decision: it was one of 8,276 separate consents required for HS2, the high-speed railway being built between London and Birmingham. So here is the puzzle: how does a process like that produce a result like this?

Bechstein's bats genuinely are rare and genuinely are protected by law. A railway genuinely does need to cross their woodland. Mitigating that impact genuinely is the right instinct — HS2's own ecologists identified the need back in 2013, long before anyone was arguing about the cost. Nobody lied, nobody acted in bad faith, and the outcome — this design, at this cost — is still one that almost everyone, regardless of how they feel about bats, railways, or HS2 generally, would call wrong.

That gap — between decisions that are individually defensible and outcomes that collectively aren't — is the subject of this piece. It shows up in the choices that produced the bat tunnel. It shows up again, in a different shape, in New York, over solar farms — and both are examined in what follows.

The greater good is not the greatest good

Each of the 8,276 HS2 consents was the product of careful work against the question it was actually asked to answer — something like: does this specific mitigation, for this specific impact, on this specific population, serve the goal it's meant to serve? For the Sheephouse Wood bat tunnel, the honest answer to that question was almost certainly yes. Bechstein's bats are rare, the woodland matters to them, and a structure that lets them cross the line safely is, on its own terms, a genuinely good thing to build.

Call that the greater good question: judged on its own, is this project worth doing? It's a comparison between two options — this, or not this — and for almost any environmental mitigation genuinely targeted at a real impact on a real population, the answer will be yes. That's not a flaw in the question. It's just that "yes, this is worth doing" says nothing about how this allocation compares to everything else — HS2's other 8,275 consents, or biodiversity spending generally — drawing on the same justification.

The question that would say something about that is different in kind, and has to be asked at a different level. Call it the greatest good question: given everything else competing for the same kind of justification — the same budget, the same goodwill, the same finite tolerance for cost or disruption — is this the best use of this particular allocation? That question is superlative, not comparative.

Critically, it can only be answered by some body positioned to see the whole set — and across HS2's 8,276 consents, nobody ever was. The scrutiny the bat structure got was about its cost in isolation. Nobody asked whether that was the best use HS2 could have made of £100 million.

This is why the greater good question, asked one decision at a time, essentially never produces a "no" for a genuinely worthwhile goal: "is this good" has no ceiling. The greatest good question is the one with a ceiling, precisely because it's relative to everything else drawing on the same justification — and it's the question that 8,276 individually honest "yeses" never collectively faced.

There's a second thing such a ceiling — an agreed limit on the whole — would do, and it points in the opposite direction from that considered so far. Right now, every individual decision is judged against no defined limit at all — which doesn't just let costs accumulate unnoticed on one side of the ledger. It also means a project that has, in aggregate, already done its fair share within some agreed budget has nothing to point to when the next objection arrives, because there's no "fair share" or "agreed budget" for it to have done. A ledger isn't only a way of saying "this is how much we've already spent, and it's too much." It's also a way of saying "this is how much we agreed to spend, and this is within it" — an answer that almost nothing currently has access to, on either side of any of these fights.

The pattern has a name. Economist Alfred Kahn called it the "tyranny of small decisions" in 1966, coined while watching passenger rail quietly disappear from Ithaca, New York, as each traveller's individually sensible choice made the service a little less viable for the next one. Ecologist William Odum extended the idea specifically to environmental degradation in 1982. Both papers are over forty years old, and the pattern they describe is, if anything, easier to produce now than it was then.

Whether £100 million was the right figure for Sheephouse Wood isn't a question this piece needs to settle. What's worth noticing is that it could be argued about at all. The structure was large enough, singular enough, to produce a number, a regulatory statement, a ministerial soundbite — a public controversy, with figures attached. Most decisions of this shape never get that — not because they matter less, but because no single one is large enough to force the issue into the open: one of thousands, each too small to notice on its own.

And that debate has already produced a response.

What happened next

The bat tunnel didn't stay a curiosity for long. As the Planning and Infrastructure Act passed through Parliament, becoming law in December 2025, ministers cited it at least ten times to make the case for reform. Among its provisions is a new Nature Restoration Fund. Instead of each project working out its own mitigation in isolation, Natural England — the same body from the bat tunnel — can draw up an Environmental Delivery Plan (EDP) covering a protected site or species. An EDP sets a maximum level of development that site or species can absorb. Developers can then either carry out their own mitigation, project by project, or pay into a levy that funds conservation for it. A ceiling on cumulative impact: on paper, that's close to exactly the kind of accounting the rest of this piece is about to argue is missing. Whether it's the real thing or the old gap with a new name on it is fiercely disputed — conservation groups call it "pay-to-destroy"; developers worry it won't be built out fast or broadly enough to matter. Either way, the argument is far from over.

Meanwhile, there's an almost-analogous situation playing out in New York — over solar farms rather than trains and bats. Looking at both means asking the same underlying question twice, at two different stages: once about a mechanism that exists and is being fought over right now, once about a problem that doesn't have a mechanism yet at all. And once you've seen the shape of the gap itself, it's hard to stop noticing how often "nobody did anything wrong, and yet" is the actual story behind outcomes that everybody agrees are wrong.

One column of the ledger

New York's Climate Leadership and Community Protection Act, passed in 2019, gave the state two headline figures: 70% of its electricity from renewable sources by 2030, and 100% zero-emission electricity by 2040. These aren't soft aspirations — they're written into the Public Service Law, and the Public Service Commission is required to report on progress against them every two years. Whatever else is true about this side of the ledger, it is the subject of continuous, public, numeric scrutiny: how close, how far, by when.

For years, the main obstacle between "where we are" and "70% by 2030" was local permitting. A single town board could hold up a utility-scale solar or wind project for years — not necessarily for ecological reasons, but for the ordinary reasons local boards resist large unfamiliar infrastructure: noise, appearance, change. In 2020, New York created the Office of Renewable Energy Siting (ORES) specifically to remove that bottleneck. For projects of 25 MW or more, ORES can set aside local zoning it finds "unreasonably burdensome" relative to the state's climate targets, and must decide each application within a fixed one-year timeline.

It has worked, in the narrow sense it was built for. As of June 2026, ORES has permitted more than 30 large-scale projects, denying only one. Each permit carries a megawatt figure, and that figure feeds directly into the 70%-by-2030 calculation — cleanly, unambiguously, and visible to anyone who wants to check. If you want to know how close New York is to its renewable electricity target, the number exists, and it updates with every permit ORES grants.

The column that's missing

Here is what happens, on that very same permit, to the other side of the ledger.

Every ORES application includes an agricultural resources exhibit — required, detailed, specific to that project. It records how much of the site sits on what New York's Department of Agriculture and Markets (DAM) calls "Prime Farmland" — Mineral Soil Groups 1 through 4 under the state's own agricultural land classification system — and DAM has a guideline to go with it: no more than 10% of a project's area should fall into those categories.

Nobody at the state level adds these figures up across projects, though.

The closest thing that exists is a running tally kept by Roger Caiazza, an environmental analyst who blogs as the Pragmatic Environmentalist of New York, built by going through each permit's agricultural exhibit by hand. His count at the time of writing: of 25 approved facilities with published data, only 12 meet the 10% guideline (two with no prime-farmland impact at all). Across the rest, the cumulative total now stands at 6,650 acres of "Prime Farmland" converted to solar.

New York has, to its credit, started building something that points at this gap: the New York State Energy Research and Development Authority (NYSERDA) is developing a "Smart Solar Siting Scorecard," intended to weigh exactly these kinds of siting trade-offs. But by Caiazza's reading, it doesn't appear a project can actually fail it — and even finished, it would score one project at a time. It would tell you whether this site was sited well. It wouldn't tell you what New York has spent, in total, to get here.

This isn't a story about anyone hiding anything. The agricultural exhibits are public, filed with every application, available to download. What's missing isn't information, and — per the Scorecard — it isn't even effort. It's the step where somebody adds it up.

These particular figures — 12 of 25, 6,650 acres — are a snapshot, taken in 2026. By the time anyone reads this, they'll have moved, probably upward, and that's fine: the point was never the specific numbers. The point is the asymmetry between them: one of these is written into law, reported to the legislature on a schedule, impossible to dispute without disputing the statute itself. The other exists because one person, on their own initiative, decided to build it. The remainder of this article addresses whether that difference — official versus improvised — actually matters, and what it would take to close it.

Neither a side nor a motive

Most people arguing about infrastructure that carries an environmental cost — a railway, a solar farm — fall somewhere on a spectrum with two poles. At one end, development comes first, and environmental objections are obstacles to be managed or designed around. At the other, protection comes first, and development proposals are threats to be resisted on principle.

Somewhere between those poles sits what's sometimes called pragmatic environmentalism — the position that development and protection both matter, and the job is weighing them against each other with numbers rather than slogans. It's not a coincidence that the farmland tally above came from someone who explicitly claims that label. On its own terms, pragmatic environmentalism is close to this piece's argument — close enough that it can sound like the same thing, stated as a creed rather than a diagnosis.

But a creed isn't a mechanism. A running total can be published — Caiazza's is — and still have no route to becoming official, because no such route exists. Naming the missing ledger is not the same as supplying one. The pragmatic position, for all that it gets right, runs into the same wall as everything else here: a number can exist, and even be heard, without binding anything.

There's a second axis, and it's a different kind of question: call it sudden-onset environmentalism — concern that appears, apparently from nowhere, exactly when a project is proposed nearby, and rarely outlasts the proposal either way. It's real, and it cuts in every direction: in New York's solar fights, soil-contamination claims get made by people with no prior interest in soil, and funding-motive accusations get traded by both sides without much to back them up. None of that is this piece's argument, though, and none of it needs settling for the argument to hold. Sincerity is a question about whoever's asking. Whether the number exists, and whether it does anything once it does, is a question about the system — and it has the same answer no matter who's asking, or why.

A test, not a verdict

"The column that's missing" set out the two things worth establishing: whether the difference between an official number and an improvised one actually matters, and what closing that gap would take. Here's a way to find out — the same four questions, applied to two cases:

• Does a number exist?
• Does the number do anything?
• Who defines the unit?
• Is any body positioned to keep asking?

The Nature Restoration Fund

Start with the UK, where the response is furthest along.

Does a number exist? For EDPs, yes — by design. An Environmental Delivery Plan defines a "maximum level of development" for a protected feature: an aggregate ceiling.

Does the number do anything? This is where it gets contested. The Bat Conservation Trust's complaint isn't really that EDPs exist — it's that the Planning and Infrastructure Act that created them never references the mitigation hierarchy, the principle that impacts should be avoided and minimised before being offset, which is normally what gives a ceiling its teeth. Meanwhile Britain Remade, arguing from the opposite direction, wants the Secretary of State to gain a power to bypass the Habitats Regulations entirely — EDP or no EDP — for any mitigation judged "poor value for money." A ceiling that can be set aside when inconvenient is a ceiling in name only.

Who defines the unit? EDPs are drawn up by Natural England and approved by the Secretary of State — government bodies operating under direct pressure to approve development, defining the very features and "maximum levels" meant to constrain it. When the body setting the ceiling and the body that benefits from a higher one sit this close together, a redefinition and a loosening can look identical from the outside.

Is any body positioned to keep asking? Emphatically yes — the Bat Conservation Trust is already campaigning, publishing rebuttals, briefing the House of Lords committee stage. But "reacting loudly" and "winning" turn out to be different things: the Act passed in December 2025 regardless. Having a constituency with standing doesn't guarantee the outcome that constituency wants. It guarantees there's a fight.

New York's farmland

Now New York, several steps further back.

Does a number exist? Partially. Caiazza's tally exists, but it's one person's spreadsheet, not a published, regularly-updated, official total the way the 70%-by-2030 figure is for the other side. The ask — precedented by wetland no-net-loss accounting in the US and biodiversity net gain requirements in the UK — is for that to become official. It doesn't predetermine any outcome; a published total is compatible with "we have plenty of headroom, carry on." It just makes that a conclusion instead of a default.

Does the number do anything? Not yet. New York's Smart Solar Siting Scorecard exists, but a number existing isn't the same as it doing anything. Worth being precise about what's missing. It isn't an empowered body — ORES already makes binding permitting decisions, incorporating project-specific detail, all the time. What it doesn't have is a remit that includes "given everything already approved." The ask isn't a new institution; it's an existing authority's checklist gaining one cumulative line item.

Who defines the unit? The 6,650 acres cited earlier is built on New York's Mineral Soil Groups 1-4 — a state agricultural-assessment classification that New York's Department of Agriculture and Markets itself describes as "generally Prime Farmland." The USDA has its own, separately-defined national Prime Farmland designation, which doesn't necessarily cover the same acres. Same three words, two different definitions, both currently in use — which means a future "fix" could shrink the headline number simply by switching which "Prime Farmland" gets counted, without a single acre of land changing.

Is any body positioned to keep asking? New York isn't at zero. State legislators have formally written to ORES asking for comprehensive data on projects affecting protected lands and farmland — the asking stage. New York has a number (Caiazza's) and people asking for an official one. What it doesn't have yet is a constituency loud enough that a number, once official, would cost someone something to ignore.

One gap, two timetables

Put side by side, the two cases look less like a contrast than two frames of the same film. New York is roughly where the UK was before the bat tunnel became a number anyone could argue about: a real gap, a guideline nobody's enforcing, no forum yet. The UK is a year or two further on — a number exists, in EDPs' "maximum levels," and the fight over whether that number means anything is happening in real time, in Parliament. Whether that fight ends with a ceiling that holds or a ceiling that turns out to be decorative is, as of this writing, genuinely unknown. If New York's gap ever produces its own version of an EDP, this is roughly the argument it should expect to have next.

Reckon regional

Go back to where this piece started. Earlier, the goal was put as a balance: not so cautious that nothing gets built, not so generous that the fields are gone before anyone thinks to count them. But "balance" is too comfortable a word. Sometimes the right answer is that nothing should be built; sometimes it's that the fields don't matter and the project should go ahead unimpeded. The point was never to settle in the middle — it's that some body has to decide where on that spectrum the line belongs, with the whole region in view, and might justifiably draw it at either end. That is the greatest good question. The third clause gives it a name people might actually repeat: reckon regional. Different register, same missing step.

The two cases this piece has followed show that step missing in opposite directions:

• Developers frustrated by mitigation that balloons without limit — the bat tunnel is what happens when nothing limits what a single piece of mitigation can cost: £100 million spent on one bat crossing, with nobody positioned to weigh it against anything else competing for the same money.
• Conservationists watching a landscape disappear one permit at a time — New York's farmland is what happens when nothing limits what a single approval can take: 6,650 acres gone, with the only running total built by hand by someone with no authority to act on it.

Which is why "reckon regional" isn't a slogan for either side of the development-and-protection argument. The ledger provides a published, official total that gives a developer and a conservationist the same numbers to work from: how much of the "agreed budget" remains, how much a given project would use, whether the region's already done its "fair share" or still has room. Some body, positioned to see the whole regional picture, does the sum — before it's too late for the answer to matter either way.