Many of the people who died in the destruction of Pompeii were very poor and may have been slaves ordered by masters to stay behind and guard large townhouses, a study of coin data suggests.
The Roman city was buried in an eruption of Vesuvius in 79AD but previous research indicates that most inhabitants escaped in over 17 hours between its start and the pyroclastic flows of ash, lava blocks, pumice and noxious gases that overwhelmed those who remained.
The eyewitness Pliny the Younger described a “black and dreadful” cloud rising above Vesuvius in this interval, and cinders, pumice and burning rock raining down on panic-stricken crowds.

Now a study of the coins found on around 200 of the 1,150 bodies excavated at the site — which was preserved under ash and rock as if frozen in time — offers new clues to the victims’ identities and the reason for their deaths.
Archaeologist Kimberly Bowes, professor of classical studies at the University of Pennsylvania School of Arts and Sciences, studied groups of coins found alongside bodies, as well as coins in savings boxes, for an article on Roman savings habits in the Journal of Roman Archaeology.
The dataset, drawn from the publication of a detailed inventory of Pompeiian coins finds, the Rinvenimenti monetali a Pompei, comprised 431 different hoards: 206 from skeletons and 225 from savings boxes in homes and shops, with a total of 25,430 coins.
Who would be asked to stay behind over the 17 hours that it took for Pompeii to be destroyed? You asked your slaves
She reckoned that the coins found on individuals would be a good proxy for their liquid savings, as it was likely they would have sought to gather all their portable wealth ahead of fleeing, or in case they later had to flee.
Bowes found that most people were carrying sums of money that were “astonishingly low” and there was a large gulf between this majority and a minority carrying significantly larger sums. Nevertheless, even these sums were trivial relative to the price of assets such as property, land and slaves.
“Most of the people had very, very little. When you start looking at where these very poor people are found, it’s surprising, because they’re found in the largest houses. That led me to speculate that these were enslaved people, which makes a lot of sense, given who would be asked to stay behind over the 17 hours that it took for Pompeii to be destroyed? You asked your slaves to stay behind.”

She said the money found on the bodies of probable slaves could have been from small allowances given by their masters. Almost half of the hoards associated with individual bodies amounted to 1-10 sestertii, while a further 20 per cent were of 11-50 sestertii and fewer than one fifth were over 150 sestertii.
According to Bowes, the average of 5.7 sestertii found in the 1-10 sestertii hoards would have covered only one or two days’ food expenses for free citizens or up to two weeks of subsistence food costs for a slave.
She added that experts had believed that those who died at Pompeii represented a broad cross-section of the population. This was partly due to skeletal evidence, suggesting that they were reasonably well nourished. “We have to ask ourselves, would domestic slaves, of the kind who, would be working in a place like Pompeii, really have demonstrable skeletal issues?”
In fact, she said, the duties of house slaves would not necessarily have involved the sort of heavy labour carried out by agricultural slaves, and it was in the interests of slaveowners that domestic slaves were healthy and presentable.
She added that physical anthropologists had tended to believe those left behind were physically disabled or old and immobile. “I’m not convinced of this simply because of the time of the eruption. Eighteen hours is a long time to get out, so physical mobility shouldn’t be as important as, say, social and legal inequality.”
Individuals and small groups have previously been identified as probable slaves at specific sites such as “Villa B” at Oplontis in the Pompeii suburbs due to the low value of their coins and other possessions and the fact that they were huddled apart from individuals carrying greater wealth.
It is also believed that the slave population across the region was notably high. For example, one estimate puts the proportion of slaves at Herculaneum, near Pompeii, at 40-45 per cent.
Bowes said that if many of those carrying small sums, and by extension those carrying no money, were indeed slaves, it suggests that their owners, who fled the city, expected to return. “Slaves were not in any way disposable. They were very expensive. And people looked at them as an investment.”
Slaves left behind at Pompeii may have been tasked with guarding against looters or putting out fires.
As for the fact that even those carrying larger sums were carrying relatively modest amounts, she said they may have included slaves with more savings, or free Pompeiians whose money was mostly tied up in property or the banking and credit-lending system.
The total sums found in savings boxes were also generally small and Bowes said they indicated that most businesses were only scraping by, while a minority, in the best locations, prospered.
Explaining the genesis of the research, she said: “I’m in the process of writing a book on what I’m calling the economic history of the 90 per cent in the Roman Empire. So it’s a bottom up economic history of working people. And I’ve been interested in collecting metrics about these people.
“The savings study came out of a desire to understand this aspect that’s so critical to modern populations. I think you’re having the same conversation in England right now, as we are in America, about how people have no liquid reserves to meet emergencies. My question was, what does that look like in the Roman world?”