Ancient World, Demography of

Updated About content Print Article Share Article
views updated


The study of classical antiquity traditionally has been concerned with the history of Greek and Roman civilization and its sphere of influence in the Mediterranean and beyond from the early first millennium b.c.e. to the seventh century c.e. and covers sociopolitical formations that range from Greek city-states to the Roman Empire. Owing to the paucity of quantifiable evidence, demographic conditions in the ancient world are at best sporadically documented and can be reconstructed only in the most basic terms. Tombstone inscriptions, skeletal remains, and literary accounts are the most widely available sources of demographic information. In addition several hundred census returns, birth certificates, and death declarations from Greco-Roman Egypt have survived on papyrus scrolls.


Ancient populations were characterized by high birth rates and death rates. Mean life expectancy at birth is conventionally put in a range from 20 to 30 years, although these limits may have differed in particularly hazardous (e.g., malarial) or healthy (e.g., high-altitude) environments. Age records from some 300 census returns from Roman Egypt (first to third centuries c.e.) have been used to reconstruct an age distribution that is consistent with model life tables that suggest a mean life expectancy at birth of around 22 to 25 years. Corroborating evidence has been derived from select cemetery populations, a Roman schedule used to calculate annuities known as "Ulpian's Life Table," and tombstones in Roman North Africa. Literary texts suggest comparably low levels of life expectancy even among the Roman elite. However, demographic readings of most of these sources remain controversial: Age records in epitaphs are distorted by age and gender preferences, and skeletal samples may not mirror the age structure of actual past populations. An alternative reading of the Egyptian census data points to significant differences between urban and rural populations, with particularly high attrition rates in large cities.

Local mortality levels were determined largely by the prevalence of endemic diseases. Seasonal mortality patterns that are discernible in large samples of tombstones reveal great regional diversity but hint only vaguely at the nature of the underlying disease environment. Attested seasonal fluctuation in adult death rates is generally more pronounced than is the case in more recent pre-modern Mediterranean populations and suggests unusually high vulnerability to infectious diseases past childhood and correspondingly high levels of overall morbidity and mortality.

The principal causes of death defy quantification but are amply documented in ancient medical literature and other textual sources: Next to ubiquitous gastrointestinal diseases, malaria and tuberculosis played a major role. Malaria in particular gradually expanded in low-lying parts of Greece and Italy. Leprosy began to spread from Egypt during the Roman period, whereas cholera and louse-borne typhus seem to have been unknown. Smallpox first appeared in epidemic form (possibly in Athens in 430b.c.e. and probably throughout the Roman Empire in the 160s through 180s c.e.) but may have become endemic in late antiquity. Plague, confined to the southeastern hinterlands of the Mediterranean for most of this period, erupted in a massive pandemic in the 540s c.e.


Birth rates cannot be directly established on the basis of ancient sources; under prevailing levels of life expectancy, the total fertility rate must have averaged 5 to 6 (that is, five to six live births per woman surviving to menopause). The total marital fertility rate for Roman Egypt has been put at about 8 to 9 and was probably similar in other parts of the ancient world. According to Egyptian census returns, 95 percent of freeborn children were born in wedlock. These documents provide the only quantitative evidence of fertility. The documented maternal age distribution of childbirths implies a natural fertility regime with a peak in the twenties and continuing substantial fertility during the thirties and into the forties and no sign of parity-related birth control. In principle, early and nearly universal marriage for women must have necessitated some degree of fertility control within marriage. Mean birth intervals of three to four years in Roman Egypt corroborate this assumption. Cultural preferences for extended breast-feeding (for up to three years) may have contributed to extended birth spacing. A broad array of putative contraceptives and abortifacients were discussed in ancient literature, and at least some of the recommended chemical agents may have been efficient. In addition postpartum measures such as child exposure and infanticide appear to have been widely (although not universally) condoned and were not curbed until late antiquity. The extent to which families practiced intrusive forms of birth control, exposure, or infanticide remains unknowable and controversial.

In the literary tradition, elite families are most commonly associated with family limitation employed to preserve their socioeconomic standing. Evidence of high (i.e., male-dominated) sex ratios has been taken to indicate a high incidence of sex-specific infanticide but may only reflect biased recording practices. Whereas the best quantitative data, from Roman Egypt, indicate neither deliberate family limitation nor female infanticide, qualitative accounts for Greece and Rome raise the possibility of a more widespread application of postnatal measures in general and discrimination against female offspring in particular.


Greeks and Romans of all classes practiced strict serial monogamy. (Although socially and legally condoned forms of concubinage and other sexual relations of married men facilitated a limited degree of de facto polygyny, overt polygamy appears to have been confined to the Macedonian aristocracy.) Child marriage was not common. Anecdotal evidence for the age at first marriage in Greece points to 14 to 15 years for women and perhaps 30 years for men. Tombstone inscriptions for commoners throughout the western half of the Roman Empire suggest a median of about 20 years for women and about 30 years for men, whereas literary texts report earlier marriage among the aristocracy. The census data from Roman Egypt yield medians of 17 to 18 and 25 years for women and men, respectively. Women began marrying around age 12, and almost all had married by the late twenties; among men, two-thirds had married by age 30, and 90 percent by age 50. The documented prevalence of early female and late male marriage foreshadows the Mediterranean marriage pattern observed in recent centuries and reveals broad continuity over time.

Divorce could be initiated by either sex and normally was not stigmatized or constrained by legal or religious injunctions. Remarriage was common for men but rare for women over age 30. In Roman Egypt two-thirds of men but only one-third of women were still married at age 50. Before the rise of Christianity, celibacy did not have favorable connotations.

Marriages were commonly virilocal (residence of a married couple with the husband's family) and often entailed the transfer of bridal dowries, which are best documented for elite circles. Slaves were legally incapable of entering marriages but often formed de facto unions among themselves. The intensity of endogamy varied along a west-east gradient. Although consanguineous unions were rare in Roman culture (with exceptions among the elite), marriage of first cousins is well attested for Greece and the Levant. Occasional half-sibling unions are also known from the Greek world. Roman Egypt in the second and third centuries c.e. stands out for the almost unique and still unexplained practice of full brother-sister marriage that accounts for one-sixth of all unions known from census returns.

Household Structure

There was no term for the nuclear family. In addition to parents and children, the Greek oikos and the Roman familia or domus included other individuals under the control of the head of the household, such as coresident kin and slaves. Although literary and legal texts emphasize the social and legal inclusivity of the Roman family, funerary commemoration in the western half of the Roman Empire tends to focus on the nuclear family. By contrast, evidence from the eastern Mediterranean points to more complex households. The Egyptian census data reveal a split between conjugal and complex households (composed of extended or multiple families) of 51 percent against 26 percent in the cities and 37 percent against 43 percent in the countryside and a greater presence of lodgers and slaves in urban households. High mortality constricted family size. The only known average is 4.3 members in Roman Egypt. Adoption was a well-established practice, but its incidence is unknown except for the fact that it was common among the Roman elite.

Partible inheritance in Greece and Rome encouraged the fragmentation of estates. In Athens daughters would receive dowries in lieu of an inheritance, but under Roman law they could also formally inherit. Women generally could own property but were to varying degrees subject to supervision by their male guardians. However, the patriarchal character of ancient households envisaged in the legal tradition was in practice often qualified by high mortality and other dislocations.

Population Size

The population totals reported in ancient literature are frequently shaped by rhetorical stylization and ignorance. Many are merely symbolic figures, and reliable references are rare. In classical antiquity the Mediterranean and adjacent regions experienced significant population growth, with more rapid expansion in the west than in the more developed Near East. After prolonged growth following an initial slump, by the fourth century b.c.e. the Greek population in the Aegean and in settlements in Sicily, southern Italy, and the Black Sea area may have approximated five million, divided among up to 1,500 separate communities, most of them with no more than a few hundred or thousand citizens. Athens, the largest and best-known Greek city-state, had an adult male citizen population of around 25,000 to 40,000 and a total population of perhaps 150,000 to 250,000 residents, including aliens and slaves. Periodic census counts from the third to the first century b.c.e. provide a rough idea of the demographic development of the Roman citizenry. A dramatic jump from 910,000 adult male citizens in 69 b.c.e. to a total of 4,063,000 in 28 b.c.e. has been interpreted as a sign of improved coverage or of a switch to the recording of all Romans instead of adult men only. Whereas the latter reading suggests an Italian population of 5 million to 6 million, the former implies a total closer to 13 million to 14 million. Although comparative evidence from later periods lends credibility to the lower estimate, which is now favored by most scholars, this issue has not been fully resolved. The gross population of the Roman Empire may have peaked at 60 million to 70 million in the second century c.e., with 50 to 60 percent residing in the European provinces and 20 to 25 percent each in western Asia and North Africa. The empire included perhaps 2,000 cities, headed by the capital city of Rome with about one million residents.

Although from the late second century c.e. on-ward recurrent epidemics may have depressed population numbers, the eastern part of the empire in particular remained densely populated into late antiquity. Only the disintegration of the western half of the empire in the fifth century and the onset of plague in the sixth century appear to have caused substantial demographic contraction. Roman levels of population density generally were reattained in Europe by the High Middle Ages, but this did not occur until the nineteenth century in Greece and the former Asian and African parts of the empire.

Population Thought

Demographic thought was poorly developed and never formally set out in theoretical terms; stereo-typical moralizing and philosophical abstraction dominate the record. In the fourth century b.c.e the Athenian philosopher Plato considered eugenics and a fixed population size essential ingredients of his model of an ideal state. In a strongly pronatalist vein, most other sources stress the desirability of large populations as an index of military strength and general vigor. As a logical corollary, the ancient rhetorical tradition is permeated by concerns about supposedly declining birth rates and the resultant demographic contraction. This anxiety also manifested itself in Roman legislation enacted by the first emperor, Augustus (27 b.c.e–14 c.e.), to encourage nuptiality and reproduction by granting privileges to prolific couples and discriminating against the celibate and the childless, particularly among the elite. Demographic thinking in early modern Europe was profoundly influenced by this pronatalist stance. In late antiquity, Christianity brought a more systematic condemnation of fertility control and novel sympathy for celibacy.

See also: Disease and History; Historical Demography; Paleodemography.


Bagnall, Roger S., and Bruce W. Frier. 1994. The Demography of Roman Egypt. Cambridge, Eng.: Cambridge University Press.

Beloch, Julius. 1886. Die Bevölkerung der Griechisch-Römischen Welt. Leipzig: Duncker & Humblot.

Brunt, Peter A. 1971. Italian Manpower 225b.c.a.d.14. Oxford: Clarendon Press.

Eyben, Emiel. 1980/1981. "Family Planning in Graeco-Roman Antiquity." Ancient Society 11/12:5–82.

Frier, Bruce W. 1994. "Natural Fertility and Family Limitation in Roman Marriage." Classical Philology 89: 318–333.

——. 2000. "Demography." In The Cambridge Ancient History, 2nd ed., Vol. 11, ed. Alan K. Bowman, Peter Garnsey, and Dominic Rath-bone. Cambridge, Eng.: Cambridge University Press.

Hansen, Mogens H. 1985. Demography and Democracy: The Number of Athenian Citizens in the Fourth Centuryb.c. Herning, Denmark: Systime.

Parkin, Tim G. 1992. Demography and Roman Society. Baltimore: Johns Hopkins University Press.

Pomeroy, Sarah B. 1997. Families in Classical and Hellenistic Greece: Representations and Realities. Oxford: Clarendon Press.

Sallares, Robert. 1991. The Ecology of the Ancient Greek World. London: Duckworth.

Saller, Richard P. 1994. Patriarchy, Property and Death in the Roman Family. Cambridge, Eng.: Cambridge University Press.

Saller, Richard P., and Brent D. Shaw. 1984. "Tombstones and Roman Family Relations in the Principate: Civilians, Soldiers, and Slaves." Journal of Roman Studies 74: 124–156.

Scheidel, Walter. 1996. Measuring Sex, Age and Death in the Roman Empire: Explorations in Ancient Demography. Ann Arbor, MI: Journal of Roman Archaeology.

——. 2001. Death on the Nile: Disease and the Demography of Roman Egypt. Leiden, Netherlands:Brill.

——. 2001. "Progress and Problems in Roman Demography." In Debating Roman Demography, ed. Walter Scheidel. Leiden, Netherlands: Brill.

——. 2001. "Roman Age Structure: Evidence and Models." Journal of Roman Studies 91: 1–26.

Walter Scheidel