The pollution in Rome’s economic success story
Roman silver production left a record in Arctic ice. The same pollution that helps reconstruct economic growth also complicates the idea of prosperity.
Some of the evidence for the Roman economy is preserved in Greenland’s ice. Lead released by mining and smelting in Europe travelled through the atmosphere and settled into layers of snow. Much later, researchers could use that record to investigate changes in metal production.
I like the sheer oddness of that connection. A material being processed for an ancient economy left evidence thousands of kilometres away, in a place its producers weren’t trying to reach. It also makes me uncomfortable with accounts of prosperity that concentrate on how much a society could produce while treating the consequences of producing it as a separate subject.
I think the costs belong in the main account. Evidence that an economy expanded is useful, but it can’t carry the additional claim that life improved for the people living in it. That second claim needs evidence of its own.
An economic record made of pollution
A 2018 study described by its Oxford collaborators reconstructed European lead emissions using Greenland ice. Much of the pollution was associated with processing lead-silver ores, linking the record to the production of a metal used in currency. The researchers found sustained high emissions during the first and second centuries CE, alongside declines associated with periods of disruption.
That makes pollution a useful clue about economic activity. It doesn’t turn an ice core into a complete measure of an economy, let alone a measure of how comfortably its inhabitants lived. The researchers are combining physical measurements with historical interpretation, and the relationship is more interesting when we preserve that qualification.
A rise in emissions can support a story about greater production while also giving us reason to investigate a cost. There’s no contradiction in that. The difficulty comes when we describe the expansion confidently and leave the cost in a footnote because it is harder to reconstruct.
The health estimate needs its qualifications
A study published in PNAS on January 6, 2025 extended this work using three Arctic ice cores, atmospheric modelling and modern relationships between airborne lead, childhood blood lead levels and cognitive effects. Its estimates suggested that pollution from Roman-era metallurgy could have produced widespread cognitive harm, including an estimated reduction of roughly 2.5 to 3 IQ points.
That is a modelled estimate, not a measurement of ancient people’s IQ. The calculation depends on assumptions about emissions, atmospheric transport and the applicability of modern exposure relationships. The authors examined alternative distributions of mining emissions because the contribution from each ancient site is unknown.
I find the proposed reach of the exposure more revealing than the apparent precision of the IQ number. The analysis suggests that rural people far from the mines could have been affected by the same production system. Living away from the industrial activity would not necessarily have put someone outside its consequences.
This doesn’t establish that lead caused Rome’s collapse, and I don’t think the story needs an explanation of imperial collapse to be compelling. A production system potentially imposing a diffuse, prolonged burden on people is consequential enough without making it responsible for everything that happened afterwards.
Uncertainty shouldn’t decide which costs we count
There is an understandable temptation to reserve the main narrative for the quantities we can reconstruct most confidently. We can describe production, trade and infrastructure, then place less certain effects on people somewhere else. Over time, that editorial choice can make the successes feel established and the costs feel optional.
The remedy isn’t to pretend that every estimate is equally reliable. It is to describe the evidence and its limits together. An uncertain burden can belong in an assessment without being assigned a falsely exact value.
I think this is also a useful discipline when evaluating new industrial infrastructure. If the case for a project includes projected output and employment, it should include an explicit account of possible burdens on the people around it. That work needs to happen while the location, design and terms can still change. Waiting for perfect evidence after a project is built leaves fewer choices about how to respond.
The Roman research can’t tell us which present-day project to approve. It does make me more cautious about allowing economic activity to stand in for prosperity. The ice preserves evidence that production happened, but deciding whose lives improved requires looking somewhere else as well.