Rio Tinto Didn't Choose to Destroy Juukan Gorge. Its Incentive Structure Did.

On 24 May 2020, Rio Tinto detonated a charge in the Pilbara and erased a rock shelter that had sheltered humans for 46,000 years. Inside the Juukan Gorge caves were a 4,000-year-old plaited-hair belt — its DNA matched to living PKKP descendants — and tools that placed continuous occupation back through the last Ice Age.¹¹ None of it survived the blast. What it cleared the way for was 8 million tonnes of high-grade iron ore — the same volume valued at roughly $104 million USD when the permit was granted in 2013, and at AUD $135 million by 2020 prices — two figures for the same ore, in different currencies and across seven years of price movement.²³ One of the oldest known sites of human habitation on the continent was traded for a number that, against Rio Tinto's iron ore business, rounds to nothing.

The story everyone tells is that a hard-charging CEO made a reckless call and got fired for it. Almost none of that is right. The CEO inherited the plan. The decision was made years earlier, by people now forgotten. And nobody got fired — they left 'by mutual agreement,' slowly, under pressure from shareholders rather than from any sense of board conviction. The blast was not a rogue act. It was the org chart working exactly as designed.

Here is the thesis, plainly: Juukan Gorge was not a leadership failure. It was an incentive-structure failure. A commodity cycle was paying record prices for volume, and a governance architecture had quietly arranged things so that heritage could never outvote a mine plan. Put those two facts together and the rock shelter never had a chance.

What an $18.8bn line item does to everything around it

Start with the magnitude, because the magnitude is the whole point. In FY2020, Rio Tinto's Iron Ore segment alone threw off $18.837bn in underlying EBITDA and $11.398bn in underlying earnings — most of a group total of $12.448bn.¹ Iron ore was not a division of Rio Tinto. It was Rio Tinto, with some copper and aluminium attached. And the cycle was accelerating: iron ore averaged $91.0 a tonne in 2020 and $132.3 in 2021, helping push group earnings up 72% to $21.4bn.¹ In a year like that, every tonne left in the ground is money a manager is, structurally, paid not to leave there. The pressure isn't a memo. It's the air the mine planner breathes.

Now set the prize against that backdrop. The ore behind Juukan Gorge was worth, at most, AUD $135 million.⁴ Against an EBITDA line of nearly nineteen billion dollars, that is roughly seven tenths of one percent — a rounding error on a rounding error. And yet the site lost. That is the tell. When something this irreplaceable is sacrificed for something this immaterial, the explanation is never 'they really needed the money.' It is that the structure had no place to register the cost. The ore showed up on a balance sheet. The 46,000 years did not.

How heritage was engineered to lose every argument

The mechanism lives in the choices nobody flagged as choices. CEO Jacques told the parliamentary inquiry that four mine-design options had been assessed — and three of them preserved Juukan Gorge.⁴ Read that again. The company had three ways to take the ore and keep the caves, and it chose the fourth. That isn't a story about scarcity or impossibility. It's a story about how the options were weighted before anyone got to the table. The site sat inside a process where 'maximum recoverable ore' was a hard objective and 'don't erase the Ice Age' was a soft consideration — and soft considerations lose to hard objectives every single time, by design.

Rio Tinto's own board admitted this in plain language. Its August 2020 review found 'a series of systemic failures,' a work culture 'too focused on business performance,' heritage 'not accorded sufficient priority in mine planning,' and — most damning — that material new information from the 2013–14 excavations, the very digs that revealed how extraordinary the site was, 'failed to trigger an internal review of mine development plans.'⁷ That last clause is the whole tragedy compressed into one sentence. The company learned the site was priceless, filed the finding, and kept the demolition on schedule. The information arrived. The structure had no slot for it.

And the people meant to object had been structurally muted long before. The Puutu Kunti Kurrama and Pinikura people had been party to negotiations over this country since 2003¹¹ and helped excavate the caves in 2009¹¹ — yet Rio Tinto's own internal communications record no awareness of PKKP objections until mid-May 2020, when the explosives were already loaded.⁶ The real picture is years of information asymmetry compounded by contractual gag clauses in the 2011 Participation Agreement that legally constrained the traditional owners from publicly objecting or seeking an emergency injunction — clauses the parliamentary inquiry specifically called out for removal.⁹¹⁰ Consent had been won under the 1972 Act, which offered no mechanism to reverse it once the ground revealed what it held.² The veto existed nowhere — not on the board, not in the law, not in the hands of the people whose history it was.

Wasn't this just one bad CEO who paid the price?

The honest objection is that accountability did arrive: Jacques was out, two senior executives with him, the chairman gone by March 2021, and roughly $3.5 million in bonuses denied.⁵ Doesn't that prove the system corrected itself? Look closer and it argues the opposite. The departures were 'by mutual agreement,' not dismissals, and Jacques stayed in the chair until a successor could be found — months after the worst heritage scandal in the company's history.⁵ What moved the board was not principle but investors, who concluded the share price was the thing now bleeding. The parliamentary inquiry, after 23 hearings over 16 months, found Rio Tinto's own internal review 'did not fully grapple with the root causes.'⁶ When the people running the post-mortem can't see the structure that produced the death, removing a few individuals doesn't fix anything. The org chart that made the call is still standing.

And the supposed legislative fix tells the same story. Western Australia passed a new Aboriginal Cultural Heritage Act in December 2021, hailed as the durable remedy. It has since been repealed and the old 1972 Act reinstated — the very framework that permitted Juukan Gorge in the first place.⁸ The reform that was meant to outlast the outrage didn't outlast the next political cycle. Nobody rebuilt the thing that broke.

Rio Tinto did not weigh 46,000 years against $135 million and conclude the money was worth more. It never weighed them at all, because it had built a machine in which only one of the two could be entered into evidence. The ore had a line. The shelter had a 'consideration.' In a year when iron ore was minting earnings faster than the company had ever seen, the consideration was always going to lose — not because anyone hated the caves, but because nobody with the power to save them was sitting where the decision was made. The most expensive thing Rio Tinto destroyed wasn't priceless heritage. It was the illusion that a value matters to a company when it has no vote.