Demographics and the Levy


So I know this is a topic that has come up a number of times but I’m wondering what, if any, consensus has arisen.

Given the populations of the various city states, there’s no way they could support a yearly levy of 1,000 people per year.

Some of the “fixes” to this problem I have heard over the years:

  1. TSR writers didn’t understand demographics at all, pulled random numbers out of their whatevers, and thought the numbers just sounded cool. I think this is more an accurate guess than a fix, but it establishes that the numbers don’t work, why, and sets the stage for the fandom to come up with something.

  2. the numbers are accurate, and the tablelands are literally on the precipice of a population crisis that was about to blow up in everyone’s faces were it not for the events of the prism pentad. This is actually very very grim, and really fits with the whole idea of the bad guys having won and Athas being a crap sack world in its last throes.

  3. double or multiply the population numbers by 10. This would go a long way towards making a yearly loss of 7,000 lives more believable when the given population of the tablelands cities is like 160,000 people. Not even considering the number of people who otherwise did from natural causes, crime, and accidents.

  4. have the levy be a cyclical thing, where each year only 1,000 people die, and the burden shifts among the city-states each year. I actually like this one a lot, as it makes the levy a nice even 1,000 and it gets spread out over the different years.

  5. the levy started out as 1,000 per city, but the cities trade it among themselves as part of treaties: maybe tyr only pays 800, while raam pays 2,300. So long as borys gets his yearly 7,000 he doesn’t care. Iirc, one of the things Tithian did to get in Borys’ good graces was offer double the levy, so there is some precedent.

  6. the numbers are right, and the cities do their best to cull nomads and tribes to make up the numbers.

So what are the community thoughts, and is there any consensus on a workable “believable” system?


Both one and three are correct.

You can prove point one by the fact that the societies described in the fluff are complex societies, but the numbers of people are indicative of tribal or clan based societies - most likely virtually everyone would be a blood relative of the SK/SQ with those numbers.

Increase the population by 10 times, as I recommended back in the day on the WotC forums. Doing that instantly resolves the dissonance between the numbers and the fluff.


One other thing - there are no great population numbers living outside the jurisdiction of the cities. Most of the arable land is under the protection of the Sorcerer Kings, and the masses of people have congregated there for food, water, and protection (even if many are slaves).

Life out in the desert is tenuous indeed. The holding capacity of the desert is limited. New slave tribes comprising of escaped slaves could not simply escape to the desert and find a niche there, they would have to displace other people or creatures.

(Robert Adducci) #4

I’ve always just thought it was #6. I like #5 too and never really thought about it but it makes sense given that Raam is significantly more populated that some of the others.

If you’re into it, come up with some reasons for the numbers to be right, but IMO, it doesn’t matter. D&D isn’t a civilization simulator, it’s a game for dramatic mythic stories. The how’s don’t really matter in cases like this, just use these situations to create dramatic events for your games.

(Stuart Lynch) #5

Or add Option #7: City population figures are accurate, but only for the urban population of that city IIRC in Medieval times you needed 10 farmer peasants to provide enough food for 1 person in a city. The verdant belts could therefore support the 10x population suggested in option #3 without requiring a radical overhaul of established canon.

Or add Option #8: City population figures are accurate but only for the non-slave population This one is on dodgier ground as I can’t recall if any published material gave a breakdown of slave/freeman/noble/templar for the cities. Essentially this would reflect the status of slaves in Athas as being property, not people.


I’ve tended to use a combination of 7 and 8 myself. I use the Veiled Alliance book numbers and run with them being adult, non-slave city dwellers only. Anyone who lives outside the walls (most nobles, non-tradesmen and the like) aren’t counted and slaves don’t count at all. Typically 10 people outside the walls, and an average of 1 slaves per 3 free persons if you want that figure.

Sorry, I suck at tables:
| City | Urban Pop | Urban Slaves | Rural Pop | Rural Slaves | Total Pop |
| Tyr | 12,000 | 0 | 160,000 | 0 | 172,000 |
| Balic | 27,000 | 9,000 | 270,000 | 90,000| 396,000 |
| Draj | 15,000 |5,000 | 150,000 |50,000 |220,000 |
| Gulg | 8,500 | 2,833 | 85,000 | 28,333 | 124,667 |
| Nibenay | 24,000 | 8,000 | 240,000 | 80,000 | 352,000 |
| Raam | 40,000 | 13,333 | 400,000 | 133,333 | 586,667 |
| Urik | 32,000 | 10,667 |320,000 |106,667 |469,333 |

Tyr’s figures are off, many were killed during Kalak’s attempt to become a dragon and the various riots after his death. All slaves were converted to freemen but at least 4,000 are dead. This could easily be higher if you so desired but it’s still the second smallest city state so I left it as is.

In my games, the levy was always every year from every city. Plenty of slaves to fill out the levy and more could be gotten from the various tribes in the wastes.

For further reference, the lowest population growth in the middle ages is listed in wiki as 9% per year. For fun, if you take just the slave numbers you end up with the following:

City|Total Slaves|9%|8%|6%|5%|4%|3%|

Even at a 3% population growth rate among the slaves, every city state (except Gulg) would have at least a small surplus of home grown slaves after the 1,000 slave per year levy.