There are no surviving census records from Classical Greece, and even if they existed, such records would be far from complete (since they likely wouldn't count women, children or slaves).* As a result we don't have nearly enough information to establish army sizes as a percentage of the whole population. Luckily that percentage wouldn't tell us very much. In Greek city-states, all male citizens aged 18-59 were liable for military service. In addition, non-citizen free men and even slaves were often drafted. In emergencies, the entire male population would march to battle. Our best estimates for population size are derived from army numbers, not the other way around.

Greek states selected men for their military expeditions in a couple of different ways, based on the different obligations of separate citizen groups. For smaller campaigns, but also for long-term overseas expeditions, they tended to pick men from the leisure class, who were formally enrolled in state records and permanently liable to serve. In Athens, such recruitment was called "from the list" - that is, from the lists of eligible rich men that were kept in each administrative district. By definition, these men were wealthy enough not to have to work for a living. They lived off the labour of others - either as landlords or by exploiting hired or unfree labour. They could afford to be absent for long campaigns, and economic activity at home was not affected.

Their numbers, though, were necessarily small. Scholars estimate perhaps 15% of the citizen population counted among the leisure class. Model life tables suggest that the military age bracket of 18-59 would include 55% of all people in an ancient subsistence economy, but the youngest and oldest of these (some two-fifths of the 55%) were practically never sent out, so that we'd be left with an army of 15% (only those rich enough) of 35% (only those of the right age) of 50% (only men) of an unknown share of the total population (since this should include foreigners and slaves too). And from these there would be additional selection for service. The resulting percentage of men under arms would be tiny compared to the population as a whole.

In practice, however, the numbers used in these expeditions were fleshed out with volunteers from lower classes, especially after Greek states began to pay warriors for military service from the mid-5th century BC. This must explain why we find the infamously poor Sokrates serving in the Athenian army that besieged Potidaia for 2 whole years - an impossible commitment for a man who would have to work to get by. Since this blurred the distinction between rich and less rich anyway, the Athenians seem to have switched by the 4th century BC to a system akin to the Spartan one, in which armies were recruited by year group. For minor expeditions, a few year classes would suffice - say, only those aged 20-24. Major expeditions could see a draft of all year groups up to 39 or even 44, which was effectively a recruitment of the entire field army of the city, since (as noted above) those aged 18-19 and 45-59 were rarely sent abroad. Payment for service made military service possible for most of these men. Their absence would certainly have disrupted the economy, at least in terms of agriculture, in which the great majority of citizens were employed; therefore, as a rule, the larger the Greek army, the greater the pressure to disband and return home as soon as possible.

The fact that there were no soldiers, and that all armies consisted of shopkeepers, farmers and estate owners, put real limitations on the ability of Greek states to keep large forces in the field. We see the pressure to force a quick outcome during the Persian Wars, when even a few weeks on campaign puts serious strain on the Greek supply system; but we see it most clearly during the Peloponnesian War, when Sparta repeatedly called on 2/3rds of its allies' total levy to invade Athenian territory each summer, but couldn't stay there for much more than a month before being compelled to turn back for the harvest. In the amateur military system of the Greeks, large armies were always temporary, and often full of people who would rather be tending their farm.

The largest armies, necessarily raised for the shorted periods, were the so-called pandemei armies - from pan-demos, "all the people". These were the population in arms. Everyone was called up to serve, from the wealthiest landowner to the lowliest beggar. The entire male citizen population aged 18-59 would be involved in these relatively unusual levies, along with the resident foreigners of the appropriate age, and all the slaves who were able to serve. These armies served for emergencies such as the defence of home territory, or for a display of power, such as the bi-annual Athenian invasion of neighbouring Megara during the early years of the Peloponnesian War. The most stupendous numbers we get for Greek armies are from these total levies, which allowed Greek states to raise forces out of all proportion to their population size. The resulting armies would be clumsy, unwieldy and full of recruits who were all but useless, and they could not be maintained for more than a few days without totally disrupting the socio-economic life of the community. They represented the ultimate effort of a city-state in its own defence.

 

*) The sole exception is the very late record of the supposed result of the census taken by Demetrios of Phaleron at Athens around 322 BC. However, the numbers do not include women or children, and they are extremely controversial, especially the estimate for slaves, which is certainly much too high.