Historoll · Episode 03 companion

Feed Rome

It is 1 November, around AD 50 (the year is only for illustration). The sea is closing for winter, and a city of about a million people must live on what is in the warehouses until the ships return in spring. You are in charge of Rome's grain. Choose a crisis, give your orders, and see if the city makes it through the next 18 months.

The crisis
Your orders
Historians' estimates run from 450,000 to 1.2 million. Ships, warehouses and stocks stay sized for a million, so this changes only how many people share them.
—
Press play
days of grain left on 1 Nov AD 50

Days of grain = grain in store ÷ what the whole city eats in a full day. It doesn't stretch with rationing.

This run 15 days: riots Sailing season

    How the simulation works

    Every day, grain ships reach port from Africa (60% of the grain; March to mid-November, short trips), Egypt (30%; one slow convoy, mid-June to October) and Sicily and elsewhere (10%). An average ship carries about 275 tons (250 tonnes). The chance of losing a ship rises outside the safe sailing season, and the sea is closed from 11 November to 9 March. Grain takes 3 days to get up the Tiber. The city eats about 1.4 lb (0.65 kg) of grain a person a day, roughly 260,000 tons (237,000 tonnes) a year for a million people.

    Days of grain is the grain in store divided by what the whole city eats in a full day. It doesn't stretch with rationing: on a three-quarter ration, 30 days of grain last 40 days.

    Warehouses. Grain that doesn't fit is sold on or never shipped, so the stock going into winter is whatever the last few summers left. Each run starts after 3 normal years. The dates (AD 50–52) are only for illustration.

    Calibration. We tuned the warehouse space and the import surplus so a normal winter falls below 15 days about once in 25 years and runs dry about once in 45. Each run spans two winters, so about 6% of normal runs dip below 15 days. Rome recorded lows of 15 days or less three times between AD 41 and 70: 7–8 days (AD 41), 15 (AD 51) and 10 (AD 70). None ended in mass death.

    Assumed numbers, stated plainly (no source gives them):

    • Stock on 1 November: comes out of the model at about 5½–8 months of grain.
    • Warehouse space: 240 days of a million people's grain, about 172,000 tons (156,000 tonnes); doubled, 480 days.
    • Imports: a normal year brings 8% more than a million people eat, before shipwrecks.
    • Harvest swings: each province's harvest varies by about 15% from year to year (one standard deviation).
    • Ships lost: 1% of voyages in the safe season, 4% in risky months, 12% in winter.
    • Riots: below 15 days of grain, a 3% chance of a riot each day (AD 51: a crowd mobbed Claudius with about 15 days left).
    • Orders: the three-quarter ration, the 1 in 10 sent away, and how much grain winter sailing moves into winter (a quarter of the summer rate from Africa).
    • Crises: Egypt misses one season; Africa's revolt lasts 9 months; the storm comes in summer, with a third of the 300 ships carrying grain.
    • Deaths (modern-famine comparison, not Roman data). Nobody dies while the city is less than 30% short of a full ration, so a three-quarter ration alone kills no one. Deaths begin only after about 3 months of deeper shortage and reach full strength at 5 months, as in Leningrad (1941–42) and the Dutch Hunger Winter (1944–45). With no grain at all, 4% of the city dies each month, Leningrad's worst. August to October is worse (Rome's late-summer disease season), but never above 4% a month. Treat deaths as a rough guide.

    Sources