Goodhart's Law: When Conservation Metrics and GDP Hide the Truth

1 week ago

I've mentioned before that I'm not a huge fan of Britain's managed moorland. To me it doesn't feel natural or wild, it doesn't even feel like a functional ecosystem. It feels sterile, like a timber plantation. Some wildlife does make a home here though, in between the grouse shoots, the burning and the grazing animals we find species of ground nesting birds like the curlew and lapwing, both of which are red listed in the UK and classified as Near Threatened globally due to catastrophic declines.

That these species are counted and monitored then is no surprise. It is always important to track declining species and be able to judge which interventions work to improve their numbers. Conservation work should be data led, our natural world is far too depleted to just make a guess and hope for the best.

So with this in mind, what do we make of the shooting industry regularly pointing to studies like the Game & Wildlife Conservation Trust's work at Otterburn, which showed curlew, lapwing, and golden plover breeding success improved "up to three times" when gamekeepers controlled predators - or to be more specific when they trapped and killed foxes, crows, weasels and stoats. Does this mean those managed moorlands are actually excellent habitats? Thriving ecosystems supporting a diverse array of species? A complex web of natural relationships and trophic cascades? Well, the numbers are real, the specific monitored species do indeed increase. The studies are peer-reviewed,  Moorland associations cite them constantly as proof they're "conservation heroes.” But strip away the conservation language and what do they show? That killing predators increases prey numbers. What a revelation.

This sort of study says nothing about the health of the environment itself and yet the Otterburn study and others like it are used by the game industry to support their claim that they are the ideal stewards of the moorland environment, that they know best and that their management techniques of predator control, controlled burning and heavy grazing are exactly what is needed.  The fact that those techniques coincidentally create the ideal environment for raising, releasing and shooting millions of game birds is just good luck on their part I'm sure.

Meanwhile, on those same moorlands, legally protected raptors are being systematically killed. Research shows hen harriers are ten times more likely to die or disappear on grouse moors than elsewhere. Thirty-one percent of satellite-tagged golden eagles in Scotland disappeared under suspicious circumstances, almost exclusively near grouse moors. The UK should have around 2,500 pairs of hen harriers based on suitable habitat. We have roughly 600. We're missing nearly 2,000 potential breeding pairs.

So the curlew numbers look great while the raptors have been exterminated. In fact even worse, an independent study has shown that up to half of all animals killed in the name of moorland management are not even the target species - they are hedgehogs, small birds and other non predatory animals who just happen to get caught by the traps.

What does all this mean? Well for one thing a perfectly reasonable and desirable measure of one particular thing - tracking declining species populations - has been weaponised by the game industry to promote an entirely different narrative: that industrial-scale game bird production equals conservation.

This is Goodhart's Law in action: "When a measure becomes a target, it ceases to be a good measure."
And it's an interesting parallel with Gross Domestic Product (GDP) - the measure of economic output (Ha, bet you didn’t see an economics lecture coming!)

Originally created as a simple way to track economic productivity during the Great Depression, its creator Simon Kuznets warned from the start that GDP "can scarcely be taken as the measure of the welfare of a nation." It was meant to answer a straightforward question: how much stuff are we producing?
But just like curlew counts, GDP got elevated far beyond its intended purpose. Post-World War II GDP became the scoreboard for national success. Politicians loved it because a single number was easy to communicate. Increasing GDP became the answer to everything: trade deficits, unemployment, inequality. Whatever your ailment, growth was the cure.

The problem with this is that GDP measures economic activity, not economic value. A car accident boosts GDP twice - once when the crash happens (emergency services, hospital visits), again when you repair the car. Tear down a functional building and rebuild it badly? More GDP. Pollution cleanup counts just as much as pollution creation. Financial speculation that produces nothing tangible? Counts as growth.

GDP was never meant to tell you if people's lives are improving. It was meant to count transactions. But once it became the target, we started optimizing for transactions regardless of whether they improve wellbeing—or actively harm it.
Just like the shooting estates can point to rising curlew numbers while systematically destroying the broader ecosystem, politicians can point to rising GDP while wages stagnate, homelessness increases, and environmental degradation accelerates. The metric looks good. The thing we actually care about gets worse.
In both cases, a proxy metric chosen for practical, legitimate reasons gets elevated beyond its purpose by those whose interests it serves.

So What Should We Measure Instead?

For Economics:
Economists have been developing alternatives to GDP for decades, precisely because its limitations are so obvious. The Genuine Progress Indicator (GPI) starts with the same consumption data as GDP but adjusts for reality: it subtracts environmental costs, adds the value of unpaid care work and volunteer labor, accounts for income inequality. Where GDP counts pollution twice (creation plus cleanup), GPI subtracts it as a cost.

Kate Raworth's Doughnut Economics offers a completely different framework: meeting the needs of all people within the means of the planet. The "doughnut" has an inner ring representing social foundations (healthcare, education, political voice) and an outer ring representing planetary boundaries (climate stability, biodiversity, clean water). The goal isn't endless GDP growth—it's thriving in the space between deprivation and ecological collapse.

These alternatives exist. They're scientifically credible. Governments and institutions resist them because they would expose uncomfortable truths: that "record economic growth" often means wealth extraction, that rising GDP can coincide with declining quality of life for the majority.

For Ecosystems:
Measuring genuine ecosystem health is more complex than counting curlews. Scientists have been working on this problem for years, and while there's no single perfect metric, there are far better approaches than species counts that can be gamed:

  • Trophic completeness: Are all levels of the food web present and functioning? A healthy upland ecosystem needs apex predators (which we exterminated), raptors (which grouse moors systematically kill), mesopredators at natural levels, and prey species in balance.
  • Functional diversity: Not just how many species, but what roles they play. Pollinators, decomposers, predators, seed dispersers—a functioning ecosystem needs the full cast of characters.
  • Ecosystem condition: Is the habitat expanding or contracting? Are natural processes functioning—woodland regeneration, predation, natural disturbance regimes? Can the system absorb shocks without collapsing?
  • Soil and hydrological health: On moorland specifically—peat depth and condition, water storage capacity, carbon sequestration, mycorrhizal networks.

A genuinely healthy moorland would have lynx controlling fox populations naturally, raptors and ground-nesting birds coexisting under balanced predation pressure, natural woodland regeneration in valleys, intact peat systems storing carbon. You wouldn't need gamekeepers constantly intervening because the system would regulate itself. That is literally nature in action.
The research frameworks exist. Conservation scientists know how to measure ecosystem integrity. But they require acknowledging that a landscape requiring constant management to prevent collapse isn't healthy—it's on life support.

Why Write About Moorland and GDP Together?

Because both taught me the same lesson about how metrics get weaponized.

I started researching moorland management because something felt wrong - the landscape looked sterile, the conservation claims didn't match what I was seeing. Other people I know love the “wildness” of moors but I've always just felt empty looking at them. So I looked into what British moorland really is, discovering its manmade origins, the ownership and management by large shooting estates. I found the studies focused on lapwing and curlew breeding success being used to justify predator extermination and peatland burning. A legitimate conservation metric twisted beyond recognition.

I realised that I'd seen that same idea elsewhere. Goodharts law again and again. GDP growth hiding wage stagnation. "Job creation" numbers that don't mention zero-hour contracts. "Educational achievement" metrics that don't measure learning. Every time, the same structure: a measurement chosen for practical reasons gets elevated into THE measure of success, and suddenly it's hiding more than it reveals. Sometimes this is malicious, other times it's no more than is demanded of the system we live in - charitable funding for example is often dependent on reaching certain targets, so you optimise for those targets rather than the main goal. 

So this isn't a post about some unique evil inherent in shooting estates. They didn't invent this playbook, they're just doing what powerful interests can do when given the chance: finding a metric that makes their preferred activities look beneficial, then defending that metric as if it's objective truth.

In the case of shooting estates their metrics justify maintaining land in ways that suit them, to the detriment of nature and the wider public; in the case of politicians and GDP their metric justifies them clinging to neoliberalism, either through genuine preference or a lack of courage to try anything else. GDP is increasing therefore the economy is growing therefore everything is good! Even as wages shrink, house prices become unaffordable and inequality skyrockets.

This isn't about claiming every politician is corrupt, that landowners are part of some evil conspiracy, or that all metrics are lies. Those are simplistic non-answers - ways to ignore problems by blaming shadowy elites or diving into conspiracy theories. Reality is far messier. We need metrics to make sense of complex systems and inform decisions. But when those metrics get stretched beyond their purpose, especially when it's done intentionally and causing real harm to people and the natural world,  that's when we need to push back.

Back on the Moor

Here's the uncomfortable truth: there probably isn't such a thing as a "healthy British moorland ecosystem"—at least not in the way we currently understand moorlands.

What we call moorland is mostly the result of thousands of years of deforestation, grazing, and burning. Left to natural processes, these uplands would become a complex mosaic. On well-drained slopes, scrub would give way to native woodland - birch, rowan, oak, Scots pine - with lynx and wolves controlling deer, raptors hunting freely. In steep valleys, clough woodlands would thrive. But on deep peat where water tables are high, blanket bogs would remain - naturally treeless, waterlogged habitats that have been bog for thousands of years. These bogs are vital carbon stores and natural ecosystems in their own right.

The shooting estates aren't wrong that their management supports some species. Curlews and lapwings do breed there. But they're optimizing for an artificial system that requires constant intervention to prevent it from becoming what it naturally wants to be, maintaining an arrested, simplified version of what should be a dynamic, complex landscape. Instead of a mosaic of woodland, scrub, wet heath, blanket bog, and natural grassland shaped by wild herbivores and predators, we get a monoculture of managed heather, burned on rotation, with anything that competes with grouse systematically removed.

Which makes the whole "conservation" argument even more absurd.

They're not conserving a natural ecosystem. They've created vast outdoor enclosures - artificial habitats meticulously managed to suit a few select species, with anything that might threaten those species systematically eliminated. As I've written before, what makes an animal part of nature isn't just its species, but what it does to its environment - its ecological role, the relationships it maintains, the way it shapes and is shaped by the living system around it. A curlew breeding on a grouse moor, in a landscape stripped of predators and burned to maintain artificial conditions, isn't really "wild" any more than a lion in a zoo enclosure. It's a captive animal in a fancy cage that happens to cover thousands of acres.

And these sterile, heather lined cages in which this theatrical charade of conservation is played out aren't even their ideal habitat - it's where they ended up after agricultural intensification destroyed the lowland wet grasslands and rough farmland they prefer. The shooting estates are exploiting species that have been pushed to marginal habitats by human activity, using their presence to justify preventing the landscape restoration that would actually benefit them long-term.

So the better question isn't "what does a healthy moorland look like?" It's "should we even have moorlands, or should we be letting these landscapes rewild into the woodlands, bogs, and scrub that would naturally occur here?"
With lynx and wolves controlling deer and foxes. With raptors hunting freely. With natural disturbance regimes and woodland regeneration and functioning trophic cascades. With peatlands that aren't burned and degraded, storing carbon instead of releasing it.
That's what a healthy upland ecosystem looks like in Britain. And it looks nothing like a grouse moor.
But we can't have that conversation as long as shooting estates control the narrative with their cherry-picked species counts, convincing policymakers that burning peatland to boost curlew numbers is "conservation.”