The Peopling of Mexico from Origins to Revolution
Robert McCaa
preliminary draft (notes omitted from web version) for:
Richard Steckel & Michael Haines (eds.),
The Population History of North America
(Cambridge University Press).
December 8, 1997


Great triumphs and terrible tragedies mark the population history of Mexico over the millennia. The first great swell of population growth in the Mexican subcontinent began almost ten thousand years ago with the domestication of gourds, squash, corn and beans. The last started less than three-quarters of a century ago—thanks to advances in public health, food production, and mass education. The demographic dynamics of the region that today we know as the "Republic of Mexico" can be conveniently divided into four great epochs: ancient (-1519 AD), colonial (1519-1821), national (1821-1910) and modern (1910-present). These political turning points in the history of this vast region had great demographic significance and therefore should not be discarded or ignored simply because they are political. Each provoked demographic catastrophe to a greater or lesser degree, but each radically transformed basic conditions of life and death in Mexico. The first three periods—ancient, colonial, and national—are discussed in this essay.

1. Ancient Mesoamerica

The peopling of ancient Mesoamerica is one of the most complex phenomena in Mexican prehistory. Perhaps because of this it is also one of the most prolifically studied and controversial. When did humans first appear on the Mexican subcontinent? Did the emergence of agriculture spark a demographic revolution? What was the role of demographic pressure in the decay and collapse of many of the great cultural centers, such as La Venta, El Tajín, Cuicuilco, Tula, Teotihuacán, Palenque, Chichén Itzá, and elsewhere? At first contact with Europeans, were Amerindians under a Malthusian threat for exceeding the limits of the carrying capacity of the land or had they achieved instead a harmonious balance with the environment? Answers to these questions are fundamental for understanding the evolution of ancient Mexican culture, politics, society and economy.

The Asian origins of the first humans in the Americas is universally accepted by archaeologists and geneticists, but considerable disagreement persists over the date of first origins—ranging from 20 to 70,000 years ago—as well as the number of migration "waves" from Asia—and whether there was one, two, three or even more. Recent, still tentative mitochondrial DNA research suggests only two: the first some 34,000 years ago followed by a second as recently as 15,000 years ago. The dating of ancient habitational sites is also highly speculative. Human habitation at El Cedral in San Luis Potosí has been placed at 30,000 BP (before the present). Sites at Valsequillo and Tlapacoya are dated to 22,000 BP. An intensive study of the Tehuacán Valley reveals continuous human occupation from 12,000 BP.

The Tehuacán Valley site offers a fascinating, if conjectural, sequence of habitational densities from remote antiquity to the moment of European contact. From 9,000 to 7,000 years ago demographic densities in the Valley barely averaged two inhabitants per hundred square kilometers (2.2 inh./100 km2). Later, first gourd, then squash and, after several thousand years, tiny corn cobs appeared in the archaeological record—two millennia after the presence of corn pollen in the Valley of Oaxaca. Over several thousand years population densities drifted upward, increasing six-fold to 14 inh./100 km2 (5400-4300 BP). Millennia passed, and the agricultural "revolution" continued, but at a pace thousands of years slower than in the Middle East. Diffusion was a multi-millennial process in Mesoamerica, slowed by the fact that expansion was along the more challenging south-north axis rather than east-west. Mesoamerican corn and other cultigens ultimately adapted both to varying day-length as well as climate necessitated for longitudinal diffussion, but this required many centuries of experience and experiment.

Demographic conditions scarcely improved with the agricultural "revolution." In Tehuacán, twenty centuries were required for a twenty five-fold increase in population. Change occurred in "jumps and spurts" and with many false starts. As irrigation technique and practice evolved (2900-2100 BP) population growth slowly accelerated and densities expanded, from 43 inhabitants per 100 km2 (3000 BP) to 165 (2500 BP) and 1,100 (2100-1300 BP). In the final phase (1300-500 years ago), in the eight centuries preceding contact with Europeans, population densities multiplied three-fold to some 3,600 inh./100 km2, fortified towns developed, and "despotic primitive states" took root.

1.a. Population in the Central Mexican Basin over three millennia

The greatest success in Mesoamerica in terms of demographic density is found in the Central Mexican Basin. From a study of more than 3,600 habitation sites William T. Sanders and his coworkers pieced together an astonishing series of population estimates stretching over three millennia. Figure 1 depicts the long-term evolution of population sizes in the Mexican basin. To take into account error and uncertainties, smoothed curves at fifty and one hundred fifty percent bracket their point estimates. The figure is scaled logarithmically—as are most figures in this essay—to suggest relative growth and decline for each period. The graph shows the population of the Valley of Mexico expanding from fewer than 5,000 inhabitants 3,500 years ago to some 1-1.2 million in 1519. Three long cycles of growth stand out—3500-2100 BP, 1850-1250 BP, and 850-500 BP—, punctuated by two periods of decay—2100-1850 BP and 1250-850 BP. Region-wide decline is explained sometimes by exogenous factors—a cooling climate or severe seismic activity—and at others by endogenous developments or lack thereof, such as population pressure, economic decay, or political disintegration.

To place colonial and post-colonial figures in perspective, I extend Sanders’s archaeologically derived estimates, first, with a historical series he constructed from written sources for population decline in the Central Mexican Basin from 1520 to 1568, and, second, with my own estimates from a nadir in 1610, followed by recovery to 1793, 1900 and 1995. If we discount the most recent and as yet incomplete phase, the graph exhibits three growth cycles at intervals of roughly one thousand years—occurring 2500, 1500, and 500 years ago.

The graph also shows how tiny rates of change, averaging less than +/-0.4% per year over many centuries, yield substantial shifts in population size and density. Thus, the agricultural "revolution," a multi-millennial process begun 4-8,000 years ago, led to a quickening of growth in Tehuacán, Teotihuacán, Oaxaca, Patzcuaro and elsewhere, but no demographic revolution, although some archaeologists interpret the same data to mean a "demographic explosion." Even when prehistoric growth rates reached their peak in the Central Mexican Basin, just 750 years ago, the annual average scarcely attained three-quarters of one percent. Figure 1 shows that this region has had only one demographic revolution and it occurred in the twentieth-century, when the annual baby crop topped two million and growth peaked at almost three percent. This revolution is already winding down. By the middle of the next millennium, twentieth-century growth may come to resemble one of the demographic swells of the paleolithic past.

Regional trends summarize myriad local experiences, fruitfully documented in the Teotihuacán Valley study. Regional variations show how difficult it was to win the demographic lottery in ancient Mesoamerica. The many "disappearances" of ancient civilizations have provoked much speculation about causes. The pioneering bioarchaeologist Frank Saul suggests that we may be asking the wrong question about the decline of Mesoamerican cities, cultures or peoples. Saul argues that the question should be "not why they declined, but rather, how they managed to survive for so long." The bioarchaeological record reveals that Mesoamerican populations (indeed, most ancient peoples) were fragile, weakened by stress, poor nutrition, and ill-health. The old notion of strong, robust, healthy populations in Mesoamerica—a pre-Columbian paradise—is poorly supported by settlement patterns and the skeletal evidence. Ethnohistorical interpretations highlight success stories, but ethnohistorical sources still await skeptical, demographically informed scrutiny.

1.b. Stress, conditions of life and paleodemography

Physical and physiological stress seems ubiquitous in Mesoamerica, although somewhat less so than among most peoples in northern North America. Osteoarthritis (degenerative bone disease), likely due to extreme physical exertion, is present in adult skeletal remains from 5,000 years ago in the Tehuacán Valley. High rates of healed fractures, severe dental wear, and advanced osteophytosis are common in the earliest extant skeletal material. Tuberculosis and treponemal infection, forms of syphilis and yaws, date from 3,000 BP. Also common are coral-like lesions on the crania (porotic hyperostosis and cribra orbitalia), severe physiological responses to acute or chronic anemia resulting from nutritional deficiencies, extreme parasitic infestation, debilitating infection, blood loss or some combination of these. The architectural riches of Chichén Itzá contrast starkly with the physiological poverty of its population, which suffered from hard labor, illness, infection, and severe malnutrition. A tally of 752 adult Mesoamerican skeletons from the Health and Nutrition in the Americas database reveals women with higher rates of facial fractures than men (gender abuse?) and more joint disease of the wrists (repetitive stress from the arduous labor of grinding corn for tortillas?). Spines of adults of both sexes show severe degenerative wear, averaging 40% or more at Jaina, Tlatilco, Cholula and Copán (Honduras). The lesson learned from these skeletons is that where the human body was the principal mechanism for growing food, constructing buildings and moving heavy burdens the biological price was great. Hard, repetitive work exacted severe wear on Mesoamerican bodies of both sexes, particularly joints required for mobility, manipulation of objects, or bearing loads.

From Black Mesa pueblo in the arid northwest to Copán in the humid southeast, the emergence of agriculture reduced dental degeneration caused by wear and tear of consuming foraged foods, but life-threatening caries, abscesses, and tooth loss became more pronounced due to high carbohydrate corn-based diets. As populations became more sedentary, diarrhea, typhus, and region-wide famine probably became more common. With the spread of a monotonous diet of squash, corn and beans, stature declined, at least for males. Shortening stature was an adaptive response to malnutrition, undernutrition and concomitant disease levels, resulting from the adoption of a settled, neolithic way of life. These were the primary causes of regional and temporal differentials in stature. Males in the north, subsisting from hunting and gathering, averaged 165 cm with little decline over time. In the center, average stature for men in the classic period fell to 160 cm. Southward from Oaxaca, the average adult male stood at 155 cm, although along the coasts heights were greater. Female stature, averaging 145-155 cm, is more perplexing because there was little systematic variation in space or time.

Paleodemography corroborates the findings of paleopathology. Extraordinarily low life expectancy was the rule for Mesoamerican populations. Paleodemographers favor life expectancy at birth as the measure of choice, but this indicator should be discounted because only extraordinary burial practices and exceptionally thorough archaeological recovery techniques yield representative samples. At most sites too few skeletons of infants and children are recovered to be credible (Teotihuacán is an important exception), and paleodemographers’ estimates of life expectancy at birth (e0) are thereby greatly inflated. The ethnohistorian Ortiz de Montellano puts life expectancy at birth for the Aztec at 37 years, but the cited source does not, in fact, support this figure. A decidedly somber picture emerges when we examine life expectancies at older ages (see Table 1). At age 15 (e15), Mesoamerican life expectancies were extremely low, ranging from 13 to 29 additional years of life. In other words, for those surviving to age 15, death came around age 28 through 44 on average. Even the most optimistic estimates are almost one-third worse than national figures for Mexico in 1940 (when e15=43 additional years, to 58; in 1980 e15=56, to 71). Indeed the figures for prehistoric populations fall well below the worst conditions in model life tables, such as Coale and Demeny’s Region South level 1, where e15=34 (to age 49) and life expectancy at birth (e0) is only 20 years.

Non-quantitative sources support the interpretation that mortality was extremely high in Mesoamerica. The Nahua (Aztec) sculpted high morbidity in stone and structured high mortality in their language. Consider the vast Nahua pantheon to beg for divine succor from a great diversity of afflictions and illnesses. Nahuatl grammar is obsessed, indeed burdened, with mortality. Why encumber the language with a grammatical suffix indicating whether kin are dead or alive unless mortality is an ever-present concern?

Extrapolating paleodemographic estimates for Mesoamerican populations points to life expectancies at birth of 15-20 years, or annual crude birth rates as high as 67 or as low as 50. Since on the whole these paleopopulations were growing, the upper bound of the crude birth rate should be set a few points higher, at say, 55-70 births per thousand population. Students of modern populations would dismiss the upper range as impossible. Nevertheless, a simple experiment reveals that a stable population with a crude birth rate of 70 and a growth rate of 0.5% per annum corresponds to a total fertility rate of 8.8 children. This is an astonishingly high figure, yet as recently as 1990, Mexican women with no schooling who survived to menopause averaged 7.5 children. This 1990 average was attained even though marriage was delayed to around age 20 and not all women formed stable unions. If we look back into the nineteenth-century we find women attaining this record, or nearly so. 8.5 was the average total fertility rate reported for carefully documented studies of Tzeltal-speakers of Amatenango (Chiapas) and the Euromestizo elite of Mexico City.

1.c. High-pressure demographic systems

Even with life expectancy at birth (e0) as low as 16 years, a high-fertility paleopopulation could sustain a growth rate of 0.5% per year. The age structure would be young, with 40% of the population under 15 years of age and 90% under 50. Storey’s paleodemographic reconstruction of a barrio in the Teotihuacan urban complex is close to the high-pressure demographic scenario envisioned here. To reach an 8.8 total fertility rate (Storey places the figure at 6 for Teotihuacan) requires 26.4 years of childbearing with the first birth occurring three years after women became sexually active and birth intervals averaging 36 months. Since menopause sets in around age 40 or 45, girls would have had to marry close to the age of puberty, at say 15 years. This is exactly what we find in the earliest extant documentary evidence for the Aztecs. Child marriage, involving cohabitation, was common in ancient Mexico. A high-pressure demographic regime is consistent with the bioarchaeological and ethnohistorical evidence for central Mexico before the invasion by Hernán Cortes and his Christian comrades in 1519.

The custom of child marriage among the indigenes surprised Europeans. Viceroy Martin Enriquez’s observation, written in 1577, is typical: "being the custom in the time of their paganism to marry almost at birth because no girl reached the age of twelve without marrying." Pictorial life histories in the Codex Mendoza depict marriage occurring at age fifteen and babies being weaned at age three (rationed to one-half tortilla [per meal?], rising to one whole tortilla at age four, and two from age thirteen). In "natural fertility" populations weaning facilitates ovulation and conception—when not actually precipitated by the birth of a second baby to be suckled. In the 1530s and 1540s, for rural Nahuas in Huitzillan and Quauhchichinollan (located in the modern state of Morelos) average age at marriage (defined as co-residing couples) is estimated at 12.7 years for females and 19.4 for males. Data for this population of 2,500 ordinary folk reflect authentically indigenous practices because the Christian spiritual conquest had scarcely begun according to the earliest surviving censuses from this region. Only one Catholic marriage versus almost 800 native unions appears in these remarkable listings written on fig-bark "paper" in Nahuatl by native scribes. These documents display an obsession with fertility, or better infertility, with scribes noting not only the indigeneous names and ages of offspring but also, for each childless couple, the number of years of marriage. The ancient Nahuas were passionate pronatalists. Sterility was a truly deadly sin, leading to the sacrifice of infecund couples, who "‘served only to occupy the world and not increase it.’"

Nahua civilization, the most successful in Mesoamerica, survived, indeed thrived, by means of a high-pressure demographic system—high mortality and higher fertility with growth rates triple those of most paleopopulations, but less than one-third post-revolutionary Mexico’s pace of two or three percent per annum maintained since the 1930s. The Nahua demographic logic can be seen as the triumph of many unconscious population experiments leading to a system of reproduction which worked over the long run. The fate of most small paleopopulations was extinction, or migration, which in the archaeological record looks much the same. The loss of a reliable water supply, an outbreak of botulism, hemorrhagic fever or life-threatening diarrhea, a lengthy period of sterility or sub-fecundity, an unbalanced sex ratio, the exhaustion of food resources—paleodemographic roulette was unforgiving.

Agriculture improved the odds of winning, allowed for greater demographic densities, and led to the emergence of towns, cities, and city-states. Urban growth meant higher mortality and greater migration to replenish urban demographic sumps (such as in Teotihuacán, as noted above), but in towns or cities opportunities for coupling also increased, thanks to a greater pool of potential mates. The Malthusian threat was not the inevitable outcome. Although by 1500, demographic densities around Lake Patzcuaro probably exceeded the long-term carrying capacity of the area, in the Central Basin technological innovations, the expansion of highly productive, raised-bed (chinampa) agriculture, improved grain transport and storage, and even warfare provided relief from the Malthusian menace. From 1519, with the intrusion of European aliens, catastrophe ensued with the death of millions from disease, exploitation, environmental degradation, and, to a much lesser extent, warfare.

2. Colonial Mexico

How many people lived in "Mexico" (central and northern Mesoamerica) when Europeans first invaded in 1519? How large was the ensuing demographic disaster, and what were its principal causes? What were the effects of Spanish conquest and colonization on Mesoamericans, on the quality of life, family, and settlement patterns? What was the demographic legacy of European colonialism? Then, with independence, did demographic decay set in or was the nineteenth-century a period of accelerated growth? Answers to these questions remain contentious, notwithstanding centuries of research, writing and debate. Now there are signs that consensus is emerging on some of these questions, in turn stimulating new insight and dialogue.

2.a. The demographic disaster of conquest and colonization

There is consensus that the sixteenth-century was a demographic disaster for Mesoamericans. Table 2 displays ten authoritative estimates of population decline for the native population of "Mexico" (or diverse parts thereof) during the first century of Spanish conquest and colonization. Estimates of the magnitude of the disaster ranges from less than twenty-five percent to more than ninety. Three schools or interpretations cluster along this broad band of figures: catastrophists, moderates and minimalists. Catastrophists place the scale of demographic disaster at 90% or more and descry a large native population at contact, exceeding ten, twenty or even thirty million. Moderates detect decreases of "only" 50-85%—disasters nonetheless. They favor smaller populations at contact (5-10 million) but agree with catastrophists on population totals at nadir (1-1.5 million between 1600 and 1650). Minimalists perceive the scale of the disaster as much smaller, on the order of 25%. The principal proponent of the minimalist position, the Argentine linguist Angel Rosenblat, is the catastrophists’ most determined critic. Rosenblat sees a decline of the native population from 4.5 to 3.4 million inhabitants, or 24%, and stabilization beginning within a half century of initial contact with Europeans. It seems to me that the population of central Mexico at contact must have been no less than the minimalist estimate of four or five million and was likely double and possibly even triple that figure.

The "war over numbers" continues because population estimates prior to 1895, when the first national census was conducted, are unavoidably crude for any large region of the Mexican subcontinent. For the sixteenth-century, the data are dreadfully crude: often derived from gross tax allotments, not actual receipts, or numbers of taxpayers, not total population. Methods for working these data are more numerical than demographic, and at best the results point to orders of magnitude. The fact remains that most places extant in 1519 were never enumerated by either native or colonial authorities. Yet today there survives a surprisingly large corpus of population-like numbers for an exceedingly diverse array of administrative units: hamlets, barrios, subject boroughs, towns, district capitals (cabeceras), and provinces. Some places ceased to exist within decades of first contact, others changed names, and not a few were relocated through the Spanish policy of congregación. Most native capital "cities," with populations ranging from 10,000 to a disputed 350,000 for the Mexica capital Tenochtitlan, survived the conquest and subsequent demographic catastrophe.

Most natives resided in a dispersed pattern of settlement, to be near corn fields (milpas), following the rules of ecology or agronomy rather than political geography. After conquest, successive congregaciones attempted to reduce natives to settlements conducive to Spanish political, economic and religious control. By 1650 wherever these efforts were successful, milpa dwellers, formerly clustered near corn fields, were forced into Spanish-style hamlets, villages and towns. Then, when authorities relaxed their grip, many natives drifted back to the fields. Nevertheless, village settlements in twentieth-century Mexico, with housing clustered around a central plaza, generally reflect colonial rather than prehispanic origins.

Sporadic censuses—and there are a few remarkably detailed enumerations from the sixteenth-century that still survive—or tax surveys of small areas capture only a fraction of this movement and are simply inadequate for estimating population totals for large areas. Baptism and burial registers, which might fill the gap, do not become available in quantity until the late seventeenth-century. The paucity of evidence has spawned much research and controversy.

The catastrophist position is best represented by Sherburne F. Cook and Woodrow Borah, the most tenacious researchers and prolific writers in the field of Mexican population history. Their point estimate of 25.2 million inhabitants in 1519 has become a talisman for many historians, while their more prudent range estimate of 18-30 million goes largely ignored. This range takes into account just two of the many sources of variation on which their estimates depend: a spectrum of average family sizes—from 3.6 to 5.0 individuals per family—and alternative frequencies of tribute collection—4 or 4.5 times per year. Cook and Borah scoured libraries and archives in Mexico, Spain, and the United States to develop the largest database of colonial population figures extant for "central Mexico," a region of one-half million square kilometers bounded in the north and west by a line connecting Tampico and Tepic and in the south and east by the Isthmus of Tehuantepec. Relentlessly quantitative, Cook and Borah standardized and converted taxes—from such diverse units as corn, cotton, turkeys, blankets and the like—into tax-payers (tributarios, casados), then into total population—with whopping adjustments for tax-exempt classes, tax-free towns, omissions, errors, and lost records. They concluded that their research documented a demographic catastrophe, "one of the worst in the history of humanity." Their point estimates show the native population imploding from 25.2 million in 1519 to 6.3 million by 1545, 2.5 in 1570, and bottoming out at 1.2 million in 1620. Their reconstruction is widely accepted, indeed, it has become a paradigm to describe the devastation of European conquests elsewhere in the Americas and Oceania. Cook and Borah’s depopulation ratios of 10-25:1 (suggesting population losses of 90-96%) are in general agreement with independent estimates by Charles Gibson, Peter Gerhard, Thomas Whitmore and others—using different data and methods for the Valley of Mexico or varying subregions of Mexico, or indeed, Mesoamerica.

As Cook and Borah’s figures gradually appeared over the years, often revising sharply upward their own earlier estimates, a great dispute ensued, particularly with specialists closest to field. Prior, more moderate reconstructions by Miguel Othón de Mendizábal and George Kubler received support from independent analyses by a younger generation of scholars. Sanders and coworkers developed a formidable challenge to catastrophist methodology and conclusions. Their systematic sample of more than 3,600 archaeological sites in the Valley of Mexico point to contact populations half those proposed by Cook and Borah for the same region, yet it must be noted that the archaeological reconstruction (projected to 1595 in Table 2) sustains the thesis of enormous demographic disaster for the native population. The statistician Rudolph Zambardino questioned Cook and Borah’s numbers and methods on quantitative grounds, urging researchers to apply ranges for each conversion factor rather than relying on point estimates alone. Zambardino favors figures of 5-10 million at the beginning of the sixteenth-century and one million at the end, a demographic collapse of 80-90%.

A third position is staked out by Angel Rosenblat, who proclaims himself a "moderate," but by my reckoning he is a minimalist. He defends his text-centered reconstruction as follows:

If in fact I did derive moderate and even low figures for the 1492 population, it was not because I had intended to do so. The data I had about the Conquest allowed no other choice, unless one were to assume vast and horrible killing, which requires a macabre imagination and which I found unacceptable given the known extermination techniques of the sixteenth-century.

Unfortunately, according to Rosenblat himself, in more than three decades of writing on this subject he rarely revised a figure or an interpretation—perhaps in part because, at least in the case of Mexico, no scholar critically scrutinized Rosenblat’s reading of the sources. After a comprehensive review of Mexican population figures from historians writing in any of five European languages, the Argentine linguist developed his own series using numbers for 1570 compiled by the royal Spanish cosmographer Juan López de Velasco. Rosenblat settled on 4.5 million as a "reasonable probability" for the native population of Mexico at contact, and 3.5 million for 1570, settling to 3.4 around 1650 (then, slowly rising to 3.7 million by 1825). Unfortunately, the pattern traced by Rosenblat’s numbers for the sixteenth-century is contradicted by the narrative on which he relies. López de Velasco concludes his own assessment of population change in Mexico with the following words: "in the beginning the natives were many more in number than there were afterward, because in many provinces, where there used to be a great multitude of them, they have reached almost the point of extinction." Rosenblat, writing after historians almost unanimously espoused the thesis that epidemic disease was the principal cause for the decline of native populations, insisted on directing his attack against the Black Legend, that the "extermination" of Indians was principally due to "vast and horrible killing."

Gonzalo Aguirre-Beltrán, seemingly a champion of the minimalist camp because he embraces Rosenblat’s figures for 1519 and 1570, is in my view a moderate. Although the Mexican scholar’s figures imply "only" a 33% decline by 1595, in fact, his complete series places the nadir at 1645 with a total decrease of 70% (to 1.3 million natives), well within the moderate camp’s 50-85% range of decline.

The catastrophists’ critics’ greatest contribution is their detailed assessment of quantitative sources and methods, which emphasizes the difficulties—even the impossibility—of obtaining satisfactory estimates from tax records alone. Case studies of specific villages, such as Tomás Calvo’s history of Acatzingo, point to contact populations as little as one-fifth Cook and Borah’s estimates. Then too, population densities implied by their figures—at more than 300 inh./km2 for the Central Mexican Symbiotic Region (20,811 km2), and exceeding 1,000 inh./km2 for Mexico City and environs—are difficult to accept. The Federal District did not attain the latter until 1940 nor the state of Mexico the former until the late 1970s. The catatrophist position is further weakened by a studied refusal to reply to challenges posed by their critics. Anyone tempted to join this debate must carefully examine the works of Cook and Borah, Rosenblat, Sanders, Aguirre-Beltrán, Whitmore and others as well as a weighty bookshelf of published primary sources. In the meantime, it is clear that before the Spanish conquest the population of the Mexican subcontinent was large—certainly five million, probably ten, and perhaps fifteen, if not twenty or twenty-five million.

In any case the thesis of demographic disaster does not rest solely on numbers. The many extant narratives provide a sound foundation for a qualitative view of the scale and causes of the calamity. That the coastal and tropical regions suffered the greatest losses is widely accepted as is the thesis that the highlands had fewer fatalities. Colony-wide losses over the course of the sixteenth-century reached at least one-half, and perhaps as much as nine-tenths over wide areas. War mortality was of decidedly secondary importance, limited primarily to a few towns in the central basin—Tlaxcala, Cholula, and Tenochtitlan (Mexico City)—, Western Jalisco (the Mixtón War of the 1540s) and the sparsely settled northern frontier, where fighting continued into the nineteenth-century. Over-work, disruption of the native economy, ecological distress, and forced relocation were much more significant than war in causing the demographic disaster, but disease remains the principal explanation for most historians, just as it was four centuries ago for the first chroniclers.

2.b. Disease and recovery

There is consensus among historians that smallpox struck central Mexico in 1520, the first of a series of devastating, multi-year epidemics that erupted in the sixteenth-century. A few months before January 1, 1521, when Hernán Cortes began his third trek to Tenochtitlan, now intent on subduing the Aztec capital by siege and sword, smallpox erupted in the heartland of the most powerful empire in Mesoamerica, killing the emperor Cuitlahuatzin, many caciques and warriors, and many women and children. The epidemic was particularly severe because, unlike in Europe, where the virus was a childhood disease, in Mexico it found "virgin soil," striking entire households, adults as well as children, in one massive blow. With almost everyone ill at once, there was no one to provide food, water, or care so that many who fell ill died, not of smallpox, but of hunger, dehydration, and despair. The Franciscan Toribio de Benavente (Motolinía) in his Memoriales recounts the lethal effects of this horror: "because they all fell ill at a stroke, [the Indians] could not nurse one another, nor was there anyone to make bread [tortillas], and in many parts it happened that all the residents of a house died and in others almost no one was left..." Native historians writing in Nahuatl are more descriptive, and less analytical or synthetic. Their eye-witness accounts privilege the unadorned facts, without editorial:

Cuitlauac...ruled eighty days after the Spaniards reached Mexico. In the time of this one, it happened that a great plague came, and then many died of it everywhere in the cities. It was said that it was the smallpox, the great raising of blisters. Never once had this been seen; never had it been suffered in Mexico. Indeed, it smote the faces of everyone, so that pits and roughnesses were formed. No longer were the dead buried; they could only cast them all into the water—for in those times there was much water everywhere in Mexico. And there was a great, foul odor; the smell issued forth from the dead.

Measles hit for the first time in 1531. When smallpox returned in 1532 and 1538, mortality was lessened because many adults, now immune from having survived an earlier attack, were available to provide care to those who fell ill. A second great multi-year epidemic struck in 1545 (cocoliztli, typhus?, hemorrhagic fever?—the identification of sixteenth-century epidemics is almost as contentious as the dispute over the number of natives at contact) and a third in 1576 (matlazahuatl, perhaps typhus carried by human lice). Although a lively debate continues over which was most severe (the German scholar Hans Prem favors the first, that of 1520-21), it is clear that the effects of each were catastrophic. The founder of New World ethnohistory, Fr. Bernardino de Sahagún, writing in November 1576 while the third great epidemic of the century was underway, recalled that because of the epidemic of 1545,

when the entire population wasted away, large pueblos were left depopulated, which afterwards were never resettled. Thirty years later the pestilence which now reigns appeared, and many pueblos were depopulated, and if this business continues, and if it lasts for three or four months, as it now is, no one will remain.

Some catastrophists project these epidemics willy-nilly to encompass the length and breadth of the Americas, but the evidence for such so-called "pandemics" is thin. For example, the smallpox epidemic of 1520 is alleged to have raged north to the Great Plains, east to the Atlantic seaboard, west to the Columbia River Basin, and south through Central America down the Andes (where, the Inca Huayna Capac died, it is claimed, of smallpox in 1525) and beyond. Daniel Reff judiciously reviewed much of the evidence for northern Mexico and concluded that there is no sign of the first smallpox epidemic sweeping beyond the Tarascan-speaking peoples of north-central Mexico (states of Michoacán and Jalisco).

Lesser crises of mumps, influenza, and others vaguely described as "plague" or "sickness" also occurred, often in tandem with famine. The eighteenth-century chronicler Cayetano Cabrera y Quintero blames higher mortality among the Indians on their poverty—bad nutrition, hunger, cold, and a lack of clothing—, excessive drinking of the native intoxicant pulque, and an intense fatalism in the face of death. He chronicles seventeen major epidemics from 1544 to 1737, in addition to the smallpox epidemic of 1520-21. The twentieth-century geographer Peter Gerhard offers an even longer list, noting fourteen outbreaks for the short sixteenth-century, eleven for the seventeenth, and nine for the eighteenth. The Mexican archaeologist-historian Lourdes Márquez Morfín extends Gerhard’s roster, particularly for later centuries, and adds primary source citations. She logs three smallpox epidemics in the seventeenth-century and six in the eighteenth (1711, 1734, 1748, 1761/62, 1779/80, and 1797). In the last century of colonial rule smallpox epidemics erupted every fifteen to twenty years, with enormous loss of life. Then on November 30, 1798, Charles IV ordered a massive vaccination campaign for all the Spanish possessions. The ensuing unprecedented philanthropic odyssey commanded by Francisco Xavier de Balmis carried the vaccine throughout Spanish America and on to the Phillipines. With independence, intermittent vaccination campaigns greatly reduced mortality, although the disease was not extinguished on Mexican soil for another century and a half.

Most historians explain demographic recovery of the native population by means of natural selection or crude Darwinian evolutionism, confusing lifetime immunity with inherited genetic resistance, but there is little evidence to support this claim and much science that negates it. Smallpox mortality was much too low to play a role in human evolution, either after 1500 in the Americas or before in Europe, Africa, the Middle East, or even Asia, where it is presumed to have originated. There is no evidence that humans ever developed genetic resistance to smallpox. For example, in the city of London, after probably half a millennium of experience with the virus, more than 2,000 smallpox deaths were recorded annually from 1710 to 1800, no less than 7% of total burials in one of the most populous cities of Europe. Indeed, Europeans, confronting the horrors of the disease, were driven to extraordinary efforts, such as quarantine, inoculation with live virus and ultimately vaccination, to staunch the spread. Genetic diversity characteristic of Old World populations may have provided an advantage, but this thesis remains controversial.

Humans did learn how to provide care to smallpox victims—water to prevent dehydration, food to relieve hunger, blankets to alleviate chill, and soothing words to offer hope—instead of fleeing in horror and abandoning the ill to die untended. In the viceroyalty of New Spain, Native Americans quickly learned how to care for smallpox victims, as is attested in pictures drawn by native artists as early as 1575. Mexicans were fortunate because here the disease remained epidemic, recurring at intervals of 15-20 years, instead of endemic as in London. When smallpox did strike in Mexico, one-in-ten or twenty might die from it, as happened with the epidemic of 1779-1780 where in Mexico City alone 12,345 deaths were attributed to smallpox. What is remarkable is that four times that number fell ill, received public charity, and recovered. With the succeeding outbreak in 1797, smallpox mortality in the City was halved thanks to timely, systematic, block-by-block, person-to-person care for more than 75,000 of its residents. Meanwhile in Guanajuato, authorities pursued preventive, yet more dangerous measures, hurriedly inoculating some four-fifths of the city’s children with pus from live smallpox virus. Only one in a hundred of the inoculated died. In contrast, 28% of the some 3,000 who went untreated succumbed to the disease—many probably infected unintentionally by the inoculated. In both cities, the smallpox mortality rate was reduced to around 6%. Care was the key in the capital, and prevention in the province.

Matlazahuatl (typhus?), another of the big killers of the sixteenth-century but in this case probably of pre-Hispanic origin, did not recur with the same intensity as in 1576 until 1736-39, when it decimated much of Mexico. In the archbishopric of Puebla, for example, almost one-third of the inhabitants died from the disease, according to parish reports compiled by the archbishopric. A recurrence in 1761-62 was preceded by an outbreak of smallpox, and although less severe this crisis still ranked as one of the great terrors of the eighteenth-century. A half century later, in June 1813, while the war for independence raged in central Mexico, the last great typhus epidemic in Mexican history erupted. Within two months one-tenth of the population of Mexico City died from the disease. By 1815, the epidemic had spread as far north as the Parral mining district in Nueva Viscaya and as far south as Teopisca in Chiapas.

Recovery of the native population began, nonetheless, by the middle of the seventeenth-century according to most accounts. Rosenblat places the nadir at 3.4 million Indians around 1650, but, as noted above, Aguirre-Beltrán reckons the figure at only 1.3 million (plus 400,000 non-Indians). It is surprising that the Argentine linguist’s figures are more than double those of the Mexican anthropologist-historian even though both cite the same source, Juan Diez de la Calle.

2.c. Race, ethnicity and social transformation

Recovery was accompanied by a great mixing of peoples of different ethno-racial backgrounds. The only comprehensive figures on the subject for the entire colonial period were crafted by Aguirre-Beltrán. Figure 2 roughs out the evolution of the three principal ethnic stocks—Indian, African and European—and their intermixtures from conquest to the last decade of colonial rule. Indians always made up the overwhelming majority of the population of colonial Mexico, and people of solely African or European origin were always only minor fractions. The second largest group by the end of the sixteenth-century was the "Euromestizos," that is, Spanish-speakers of mixed Indian and European stock. Within a century of conquest Indo-mestizos (mixed stock Indian speakers) and Afro-mestizos (Spanish-speaking mixed groups with an African component) also made up a sizeable fraction of the population.

The numbers underlying this graph always add up, but they suggest no more than orders of magnitude. It seems likely, for example, that the scale of population disaster was less before 1570 than after, contrary to what the series implies. Likewise, the notion that the growth of the mixed groups was six-times greater before 1650 than after seems improbable. Then in the late eighteenth-century, the apparent ever-accelerating increase of all ethno-racial groups may be more mechanical than demographic—due to improvements in census taking (and corrections) than to increased growth rates. The Villaseñor "census" of 1742 produced valuable reports for many parishes and towns but failed to cover the entire colony. For many places this document reports numbers of families—often crudely eye-balled—rather than inhabitants. For example, the town of Guadalajara is reported as containing "eight to nine thousand families of Spaniards, mestizos and mulatos, not counting Indians…." More common is the degree of inexactitude expressed for the town of Actopan (Hidalgo state), which reports 50 Spanish families (almost all of whom would be in Aguirre-Beltrán’s terms "euromestizos"), 2,750 Indian families and 20 of other castas ("indomestizos," "afromestizos," and perhaps an African or two).

The best colony-wide census was the last, that ordered by the Viceroy Conde de Revillagigedo (1789-93) and the first to use a standard format for listing individuals by name, age, sex, occupation, race, and marital status. Nevertheless, this effort missed large expanses of New Spain. The German savant Alexander von Humboldt, from his sojourn in the colony, prepared a four volume Essai which revised the Revillagigedo figures to produce a comprehensive set of estimates, adjusted for growth to 1803. A decade later Francisco Navarro y Noriega increased the Revillagigedo numbers by 20% for under-enumeration (Humboldt favored 10%), obtained figures for districts which had not reported earlier, and estimated growth to 1810 at 25% (1.5% per annum for 17 years, using arithmetic cumulation rather than geometric).

For almost two centuries now, Navarro y Noriega’s results remain the most widely cited, yet they surely exaggerate the true population in 1810 by perhaps as much as one-fifth. This is evident in Figure 2 by the identical, steeply sloped curves for all ethno-racial groups from 1793. The graph shows that the totals for each group in 1810 were computed mechanically. Unfortunately, the royal accountant followed the method favored by Humboldt, estimating growth from parish records by subtracting burials from baptisms. Neither Navarro y Noriega nor Humboldt paid much heed to the fact that in Mexico baptisms were always more faithfully recorded than burials (until the latter decades of the nineteenth-century when civil registration undermined the religious system). Faulty logic convinced Navarro y Noriega that annual growth was 1.5% per year—probably double the actual figure, although only half what Humboldt settled upon. Thus, 5-5.5 million seems a more likely estimate for the population of Mexico in 1810 than Navarro y Noriega’s 6,122,354 or Humboldt’s 5.8 million for 1803 (and 6.5 for 1808). If historians insist on using these figures, then estimates for earlier and later years would have to be similarly corrected for errors and omissions—a forbidding challenge.

In any case, population growth in the closing decades of Spanish rule was much less than Navarro y Noriega, Humboldt or other "triumphalists" of the era surmised. There were regions of rapid growth in eighteenth-century Mexico. One of the fastest was the archbishopric of Michoacán, northwest of Mexico City. Here population increased five-fold during the eighteenth-century, but part of this growth was due to migration into the region. Then too, there was a noticeable slowing in the final decades of the century, due to successive waves of pestilence and famine. Precisely when elsewhere in America and western Europe population increase was accelerating, in Bourbon Mexico successive calamities condemned the colony to slow demographic growth.

There is consensus that demographic recovery, in addition to growth, meant transformation. Infusions of European and African stocks were slight (and predominantly male), as Figure 2 shows. If the Aguirre-Beltrán series is sound, foreign stocks peaked around 1650 with 35,000 Africans (two percent of total population), mostly slaves, and 10,000 Europeans, mainly Spanish speakers. The most dramatic change was the growth of mestizos, or people of mixed stock, which, according to Aguirre-Beltrán, constituted almost 25% of the population as early as 1650, rising to 40% in 1810. Historians agree that in colonial Mexico racial categorizations were fluid (documents usually speak of "calidad" instead of "raza"—character or reputation, instead of race), and that passing was common. Thus, the rapid growth of the mixed population was a matter of economics and sociology, but demography was also important. Among Europeans and Africans the shortage of females insured much interbreeding, if not intermarriage, with Amerindians. Then too, social identities had their advantages, for undermining as well as upholding the colonial order. The onerous head tax levied solely on Indians encouraged some to abandon the village of birth (particularly where land was scarce or made scarce by land-grabbing Spanish-speakers) for nearby haciendas or towns and adapt to a non-Indian calidad. In Michoacán, for example, over the eighteenth-century the population as a whole increased five-fold, but the native population only tripled, in part due to Indians abandoning village life to escape the head tax.

For people of African roots—perhaps 200,000 slaves were imported into Mexico over three centuries—, slavery gradually withered away. By the beginning of the eighteenth-century, free labor was too abundant—that is too cheap—for slavery to compete. Then, too, slaves helped destroy slavery, by fleeing, extracting concessions, demanding freedom, taking advantage of civil and church law, and forming communities of free people called mulatos or pardos. Afro-Mexicans with conscious identities based on kinship and community numbered more than one-half million by 1810 and "constituted the largest group of free blacks in the Western hemisphere."

Recovery of the native population may be estimated from trends of baptismal series, once parish registers achieve a degree of consistent coverage in the late seventeenth-century. Cecilia Rabell’s clever analysis of baptism trends for nine parishes shows rapid growth in the north (as high as 1.5% per year in León and one percent in San Luis de la Paz, Valladolid, Charcas and Marfil), but decelerating growth in the center. Rabell places the point of inflection, where rates of population change turn negative, at 1693 for parishes in the center, and at 1737 or 1763 elsewhere.

Burial series for ten parishes from Chihuahua to Chiapas illustrate the vast regional and ethnic volatility of year-to-year population change in eighteenth-century Mexico (Figure 3). Burial registers offer a somber view of the long eighteenth-century (1690-1820), with deaths in crisis years mushrooming four to fourteen-fold above the norm, as in 1692, 1737, 1762, 1779, 1797 and 1813. The series depict the classic workings of ancien regime demography, in which epidemics undercut the enormous growth potential of high fertility populations. Cholula, the site of one of the first, large-scale published studies in modern Mexican historical demography, is our best known example. In normal years Indian burials in the Cholula region numbered 750-1,000, but in 1737 they skyrocketed to 16,926 (upper left panel in Figure 3). The authorities responded with the ancient Spanish refrain—"huir presto, irse lexos, volver tarde" (flee swiftly, go far away, and return belatedly)—and sponsored public processions to call upon the saints for salvation. Note however that non-Indians were not as affected by epidemics and famine (lower left panel). The vast majority of deaths were Native Americans. The disaster raged through the heartland of New Spain (causing over 40,000 deaths in Mexico City alone), but only faint ripples reached the north (San José de Parral—upper right panel) or the south (lower right) such as the Chiapan community of Teopisca (1739), and went completely unnoticed in nearby Ocozocuautla.

The south had its crises as well (Fig. 3, lower right panel). Consider the epidemic of 1693. Before it struck, the parish of Teopisca numbered some 2,000 souls. Then, in 1693, 1,510 burials were recorded, a fourteen-fold increase over the average. The vicar reported that scarcely 350 people survived the devastation. A century later Teopisca still had not recovered, numbering fewer than a thousand souls. Nearby Ocozocuautla escaped much of the destruction of 1693 with "only" a four-fold increase in burials, but in 1769-70 almost half the population of the parish died, and recovery remained elusive a half century later. These stories from individual parishes challenge the uniformitarian thesis that around the middle of the seventeenth-century widespread growth of the native population resumed.

2.d. Late colonial demographic dynamics

Demography is a science of rates—of numerators and denominators, that is of demographic events, such as deaths, and populations at risk of experiencing such events. Because we lack the makings of good rates for colonial Mexico, primarily denominators, much of its demographic history remains unknown and unknowable. Only where registration is relatively accurate, and reliable population censuses available, can series of baptisms and burials be converted into birth and death rates. Everywhere in colonial Mexico many deaths went unrecorded. High fees encouraged surreptitious burial as did the widespread belief that final rites were not required for "angelitos," innocent infants and children unsullied by sin. In high-pressure demographic systems babies and young children often account for one-half or more of deaths—but not according to colonial Mexican parish registers which yield wholly implausible infant and child mortality rates.

Baptisms were more faithfully performed, but even then, not necessarily recorded. Catholic priests in the New World were often responsible for five or ten times as many souls as in the Old, and here parishioners procreated and died with what must have seemed demonic haste, in much higher proportions than in western Europe. Here, to expedite recordkeeping, priests often jotted the bare details of baptisms, burials and other sacraments on scraps of paper. Then, as time permitted, perhaps weeks, months or even years later—if at all—notes were transcribed in sacramental registries. A vicar’s sudden death could result in the omission of days, months or years of entries in a parish’s books, or, if the successor found his predecessor’s notes, in page-after-page of hastily transcribed entries written at several long sittings over a period of days, or even months.

Consequently, the "royal road to historical demography," as Rabell calls the French "family reconstitution" method, proved to be a dead end in Mexico. The family reconstitution method is a rigorous procedure for constructing and analyzing demographically valid family genealogies of ordinary people. The five conditions required for a successful journey on the royal road—stable family names, a small parish, vital events faithfully recorded week-by-week for a century or longer, low illegitimacy, and little migration—do not seem to characterize even one of the more than one thousand parishes of colonial Mexico.

Only two Mexican family reconstitution studies have been published since the method was invented over four decades ago. Both offer intriguing insights into colonial demography, but neither is entirely faithful to the method. Lacunae in the records make for truncated studies and, in the case of Thomas Calvo’s genealogies for Guadalajara, worrisome error rates. Herbert Klein’s insightful demographic history of Amatenango, Chiapas (1785-1816) reveals a high-pressure demographic system persisting through the waning years of colonial rule. In Amatenango, Tzeltal women married at very young ages, averaging 16.1 years at first union (rising to 18.6 years in 1930, and in 1990, for non-Spanish speaking women in Chiapas, 19.7 years), and nearly all married by age 20. Frequent widowhood was quickly repaired by rapid remarriage for women as well as men. Fertility was high. Birth intervals averaged 36 months, and the total fertility rate was 8.5 children. Klein concludes that early marriage and high fertility were demographic responses to an environment rich in resources which permitted unrestricted expansion, but he may have uncovered, instead, the tenacious persistence of time-tested prehispanic patterns which facilitated survival even under the most adverse circumstances.

Family reconstitution provides valuable clues about legitimate fertility, but methodological limitations do not permit measurement of extra-marital fertility. Where non-marital fertility is common but variable, as in Mexico, other methods are required. The handful of parishes in central Mexico for which crude birth rates have been computed average 52 baptisms (births) per thousand population, remarkably close to Cook and Borah’s figure of 51 for several dozen Oaxacan parishes from the same period. Child woman ratios also point to high fertility, averaging 750 children under five years of age per thousand women aged 15-44 (data are for 91 Oaxacan parishes in 1777), compared with 5-600 in late eighteenth-century England. The total fertility rate in early seventeenth-century Tacuba (1623-1630) was eight children just as it was 150 years later in San Luis de la Paz. Reliable fertility data are too sparse to construct time series or compare social or ethnic groups. Error is likely to exceed measured differences. "Natural" fertility, without the slightest hint of birth control, characterize all reliably measured populations of colonial Mexico, although Calvo claims that he finds implicit evidence of birth control among the 256 Guadalajaran families "reconstituted" for the years 1666-1730. It seems more likely to me that this analysis does not adequately take into account children lost entirely to the church registry.

While natural fertility was the rule in colonial Mexico, in practice, attained fertility fell far short of biological limits. Marriage made the difference. "Marriageways," access to stable unions including those unsanctioned by church or state, were socially constructed and differed over time and by ethnic or socio-racial group. Stable coupling was the key to regulating fertility and reproduction in New Spain.

For rural Indians three features of the "feria nupcial" (nuptial fair or marriage market) stand out from the small number of studies completed to date. Marriage occurred at an early age and was nearly universal. Most births (85-95%) occurred in stable conyugal unions. The extreme may be two rural communities in the present-day state of Morelos, Huitzillan and Quauhchichinollan, where the average age for initiating cohabitation for females in the 1530s was less than thirteen years. By age twenty-five all women were in unions or widowed. Illegitimacy was well below 5%. Narrative sources also point to extremely youthful marriage among the indigenes. The Codex Mendoza tells the life cycle of the Mexica (Aztec) people from birth through old-age (folios 57-71). The pictographs show females ready for marriage from age 13 (Figure 4). The informed exegesis of this scene by Frances Berdan and Patricia Anawalt reads:

...the daughter now has taken on another of the household’s daily domestic tasks. The gloss states that the thirteen-year-old girl is grinding [maize for] tortillas and preparing food. She kneels before the grinding stone, pulverizing the limewater-soaked maize kernels with the stone roller. . . . The girl, whose skirt now has a decorated hemline, is becoming a "grown maiden of marriageable age," whom Sahagún defines as one who grinds corn and makes atole.

A study of Acatzingo village in the archbishopric of Puebla shows female marriage age rising from an average of fourteen or fifteen years in the seventeenth-century to seventeen or eighteen by the end of the colonial era. Elderly spinsters were rare. The age gap between spouses also narrowed over time from five or six years in the sixteenth-century to only two or three years in the eighteenth. For the last century of Spanish rule, a slow, sustained rise in marriage age has been documented in Oaxaca, Guadalajara, and elsewhere.

Early marriage of Indian women left little time for prenuptial couplings. In rural areas, bastards, including "hijos naturales," "hijos de la iglesia," or "hijos de padres noconocidos," typically amounted to ten percent or less of Indian baptisms, with a surprising tendency to decline as marriage age rose toward the end of the colonial period. These patterns did not hold in the city, but only a small fraction of Indians resided in urban settlements. In the handful of colonial towns—only ten numbered 10,000 or more inhabitants in 1750—urban Indians married much later than in the countryside, probably due to the delay occasioned by migration itself. Bastardy was also more common in urban areas, particularly in the demographic blackhole that was Mexico City, where migration necessarily compensated for losses due to appalling mortality.

For peninsular-born Spaniards, whose nuptial proclivities stood at the opposite extreme from rural Indians, migration certainly delayed marriage. Male peninsulares (there were few female immigrants) married very late, often in their thirties or forties. Many never married at all, although this was no obstacle to siring more than their share of children.

"Españoles," or Mexican-born Spaniards, evolved cultural patterns that by the end of the eighteenth-century were akin to the practices of rural Andalucía in southern Spain, where women typically married in their early twenties but as many as 20% never married at all. The age gap between spouses was substantial, with husbands often four or more years older than their wives. In New Spain, bastardy rates are typically placed at 10% for "españoles" (several points greater than in Andalucía) but racial labels are particularly problematic for bastards, many of whom are listed as orphans of unknown parentage.

Intermediate between "españoles" and "indios" in terms of marriage age were mestizos and castas who were also distinguished by unique marriage patterns. Bastardy sets these groups apart. In San Luis de la Paz at the beginning of the eighteenth-century, for example, 20% of mestizo births were not legitimate, compared with 33% for castas, but only 13% for Indians and 10% for Spaniards. Nevertheless, these differences wore thin by the end of the colonial period.

Fertility in colonial Mexico was constrained by remarriage as well. Widows are perplexing to historians, who do not understand why there were so many in the Latin American past. They insist that most must be fictive widows, that is, single mothers in disguise. For population historians, widows were the escape valve for the Malthusian pressure cooker of colonial demography, a socially accepted, and religiously sanctioned, means of restraining fertility. Consider that within 15 or 20 years of first marriage most conjugal unions were disrupted by the death of a spouse, usually the male, as discussed below. Widowed males readily remarried, regardless of social or ethnic origins, but this was not the case for females, particularly non-Indians. Españolas who became widowed faced poor prospects in the nuptial fair, especially as they neared or surpassed their thirtieth "April." The full reproductive potential of Indians was least likely to be curtailed due to widowhood, because Indian widows remarried in greater proportions. (Were differences in remarriage prospects for widowed españolas and indias a matter of wheat bread versus corn tortillas, that is, a function of time necessary for preparing the daily bread?) Over the eighteenth-century, marriage and remarriage patterns converged in mixed-ethnic communities throughout New Spain, but large contrasts in nuptial practices remained between Hispanicized settlements and Indian villages and between town and countryside. The cultural and material conditioning of remarriage and marriage, family and household implied by ethno-racial differences is yet to be fully mapped for colonial Mexico.

The history of life expectancy also remains largely unwritten, but not for a lack of effort. The task is complicated by the poor quality of burial registers. The classic technique of computing mortality rates from recorded ages at death in a parish (numerator) coupled with a corresponding population census (denominator) yields implausibly low figures for colonial Mexico. At best, clues emerge from the inventive analysis of odd pieces of data: proportions orphaned or age distributions of skeletons, burials or populations (Table 3). These straws-in-the-wind all point to the brevity of life in New Spain, near or even below the worst levels recorded for early modern Europe.

Life expectancy estimates for New Spain approach, but rarely attain model life table level 1 proportions (Table 3). An exceptionally detailed census from the 1530s written in the Nahuatl reveals that one-in-seven of children aged 5-9 no longer had a surviving father. Of 261 children at that age 15.6% were paternal orphans. This figure is three percentage points worse than level one mortality (20%), suggesting a life expectancy at birth of 16 years or less. A half century later in Teotihuacán, mortality conditions were so terrible as to imply extinction (e15=13 years!)—if we accept results obtained by conventional paleodemographic methods. Written records and better methods suggest sustainable conditions but little variation over time or space (e15=29-33 years). There is evidence from the eighteenth-century that men (e15=35) lived longer than women (e15=31), at least in the north central predominantly Indian parish of San Luis de la Paz. On the northern frontier at the beginning of the nineteenth-century, only 46% of fathers of grooms were alive when sons aged less than twenty-five years married. To attain level 1 conditions (e0=20), the figure should be bumped up five percentage-points, or, to reach national levels of 1940, when life expectancy at birth was 40 years, increased by exactly one-half to 70%.

When independence was finally won in 1821, the result was a political victory, not a social or demographic one. The demographic legacy of the Paleolithic past and three centuries of colonial rule remained. The ancien regime’s high-pressure system—high fertility checked by high mortality—left a small, yet significant, margin for population increase. Indeed, Malthusianists see demographic growth, heightened poverty, inequality, and discontent spawned by an inflexible colonial order as the roots of rebellion and the struggle for independence. Enrique Florescano’s evidence of price increases for corn and wheat preceding epidemics suggests Malthus’ "positive check" at work, but other historians find colonial grain price series less persuasive and favor epidemics as the autochthonous regulator of demographic change in Bourbon Mexico. Questions of whether population growth lead to misery and social crisis, to agricultural innovation and expansion, or to social and cultural transformation, even protoindustrialization, call for further research.

Demographic revolution incipient in late-colonial Mexico would not erupt full-blown until the 1930s. In Europe, the transformation was already well underway a century earlier. By 1780, England’s death rate was probably half that of Mexico’s and its population growth rate likely twice as great. Epidemics in Bourbon Mexico were at least twice as devastating as in western Europe at the same time. Grain price fluctuations no longer governed the European mortality regime, unlike in Mexico, where in 1786, the "year of hunger," 15% of the population of Michoacán died of starvation and related causes. Indian mothers combated starvation by selling their children for a few coins (reales), the daily wage of an unskilled laborer in normal times. The first steps toward demographic revolution were taken under Spanish colonial rule: banning the use of corpses by beggars seeking alms, barring elaborate public burial for victims of epidemics, prohibiting the sale or rental of epidemic victims’ clothing, imposing quarantine to contain the spread of infectious disease, mobilizing large scale relief in times of crisis, vaccinating the mass of the population against smallpox, etc. With independence greater strides would be made, but at the same time even greater ones were expected.

3. Independent Mexico’s Censured Century

The nineteenth-century was a disappointment for many Mexicans. Independence in 1821 followed a dozen years of war, an enormous loss of life and widespread economic destruction. Then, over the ensuing half century of nation-building, Mexicans fought innumerable insurrections and civil wars, lost Texas to English speaking immigrants, found themselves at war with the United States, signed away the vast northern territories for a trifle, suffered the humiliation of invasion by a European army and, before the chaos ended, were ruled by a Hapsburg princeling under the patronage of Napoleon III and a French imperial army! Grand dreams of making independent Mexico into a prosperous, populous nation were confounded by decades of political turmoil, civil war and invasion.

Education and immigration were supposed to transform independent Mexico, but almost a century after Father Hidalgo’s "Grito" in 1810 this republic of 13.6 million people claimed barely two million citizens who could read. The literacy rate for females scarcely reached 20%, nine points less than males, and both fell far short of expectations.

Immigration was an even greater disappointment. In 1900, after decades of promoting European immigration, 99.5% of the resident population of Mexico was also born in Mexico. Resident foreigners totaled only 57,491 and came mainly from three countries—Spain, the United States, and Guatemala—hardly what the proponents of immigration had in mind. Then too, most immigrants to Mexico were male. According to the 1900 census, only sixteen thousand foreign-born females resided in the entire country, a rather meager base for remaking a race. The peopling of Mexico has always been a matter of natural increase with little immigration and the nineteenth-century was no exception.

Mexican political leaders, from the founding of the nation, equated population growth with economic prosperity and political power. As early as 1830 Salvador Piñeyro, the governor of the Republics’ most remote and poorest state, Chiapas, declared the debate regarding what constituted good or bad government as resolved. For Piñeyro, population figures settled the matter. His simple rule would categorize most nineteenth-century governments as "bad," because until 1895 none managed to conduct even a national census, notwithstanding a constitutional requirement that the population be enumerated regularly for purposes of political representation.

3.a. An alternative scenario of demographic change for nineteenth-century Mexico

Among historians, the conventional view of the nineteenth-century is a gloomy one, that demographic growth was slow well into the second half of the century, at less than 0.5% per year. Then from 1876, growth accelerated to 1.5% per annum and continued at this pace for the ensuing quarter century, when, after the passage of almost one hundred years, the supposedly high growth rates of the last decades of Spanish rule were finally regained. If the symmetry of this scenario is seductive, the evidence supporting it is not persuasive. As I have argued above, population figures for 1800-1813 were inflated because the authors of the most commonly cited numbers—Alexander von Humboldt and Francisco Navarro y Noriega—greatly exaggerated the true rates of growth in late colonial times and corrected their figures in an entirely mechanical way.

Table 4 (and Figure 5) contrasts the conventional scenario for the nineteenth-century with a series constructed for this essay. The pattern suggested here would halve the commonly accepted growth rate for late colonial times, double that for the early republic, and retain the high rates of the late nineteenth-century. I am persuaded that during the last decades of colonial rule mortality conditions were worse than during the early Republic. Before discussing the mortality evidence consider the overall pattern of growth and spatial distribution of Mexico’s population in the nineteenth-century derived from these alternative figures.

From some five million inhabitants in 1800, Mexico grew to eight million by 1855, and to over 15 million in 1910. This tripling of the population over barely one hundred years probably equaled or exceeded the record for any other period in Mexican history prior to the great demographic revolution of the twentieth-century. After 1910 the population of Mexico increased more than five-fold in nine decades, surpassing 80 million in 1990. Growth in the nineteenth-century was well below the record of the twentieth, but was substantial nonetheless.

Figure 5 compares total population figures compiled by the National Statistical Institute (INEGI) with my totals constructed from state-level data. Because of the notorious deficiencies in baptism and burial registers, the new series discards any figures derived from Humboldt’s widely used, but wrong-headed, birth-minus-death equation. Now, the 1790s appear as years of noticeably slower growth than the first decade of the nineteenth-century. The worst decades of that century were those of war: 1810, 1840 and 1860, with 1850 perhaps slightly better (a civil war erupted in 1857). The 1820s and 1830s were periods of higher than average growth, but these years did not attain the levels of the closing decades of the century.

In the early years of the twentieth-century population growth tapered off, due to a slight decline in birth rates, but largely because of Mexican emigration to the United States. The number of Mexicans enumerated north of the Río Bravo grew slowly from 68,000 in 1880 to 78,000 in 1890 and 103,000 in 1900. In 1910, the U.S. Census Bureau enumerated 221,915 Mexican born nationals, but estimated the total "Mexican race" population at 367,510, including citizens of Mexican parents born in the United States. As the economies of the southwestern United States boomed, Mexicans flocked into the region, nearly half of them to Texas and more than one-fourth to California. When revolution erupted in Mexico on November 25, 1910, almost 2.5 percent of native-born Mexicans already resided in the United States. While this was double the figure of just ten years before, it was only one-third of what it would become following a decade of revolution, civil war, banditry, famine and epidemics.

3.b. Regional change, migration, urbanization, and the persistence of Indian language speakers

The Mexican Republic’s states and regions show great diversity in patterns of demographic change over the nineteenth-century. The slowest growing regions were the south and east, encompassing Oaxaca, Chiapas, Tabasco, Campeche, and the Yucatán. This region shrank from one-fifth of Mexico’s total population at the beginning of the nineteenth-century to one-seventh at the end. In contrast, the north doubled its demographic weight from 8% to 16%, notwithstanding the mid-century amputation of Texas, New Mexico, and Upper California. In the center, political chaos, beginning with the "Grito" on September 16, 1810 and ending with the execution of Austrian Archduke Ferdinand Maximilian on June 19, 1867, provoked nearly one million war deaths and displacements. Consequently, the Federal District and nine neighboring states shrank in relative size, from 49% in 1810 to 40% in 1870, and then recovered to 43% by 1910. The center north, on the other hand, (Jalisco, including Nayarit, Colima, Guanajuato, Aguascalientes, Zacatecas, San Luis Potosí, and Querétaro) was a place of refuge, gaining population during the early decades of the century when political unrest was greatest, and losing in later decades when economic growth was more vigorous elsewhere.

Notwithstanding these disparate regional growth patterns, migration between states seems to have been relatively modest over much of the nineteenth-century. When reliable nation-wide migration data finally become available with the census of 1900 only 6.6% of males—and 6.1% of females—lived outside the "state" of birth ("federal entity" in Mexican statistical parlance). Almost one-third of all inter-state migrants resided in the national capital, Mexico City (Figure 6). Three entities accounted for almost half of all migrants: the capital, Veracruz and Coahuila. Twenty-six federal entities attracted fewer than 50,000 migrants each; eighteen, drew fewer than 25,000, accounting for 21% of all migrants. Neither sending nor receiving many migrants was four hermit states—Chiapas, Oaxaca, Yucatán and Guerrero. Fewer than 3% of their native sons and daughters lived outside the state of birth and a like proportion of current residents were in-migrants.

In half of all states migrants made up less than 4% of the population, and in only seven did they account for as much as ten percent of total population (in Aguascalientes, Coahuila, Colima, Durango, Morelos, Nuevo León, and Tamaulipas). The fact that Morelos, a seedbed for revolution in 1910, was a magnet for migrants—and the only central state whose residents withstood the seductions of the Federal District—suggests that material conditions may not have been as bad in the land of Emiliano Zapata as historians of the Mexican Revolution would have us believe.

In the Federal District itself, migrants constituted over one-half of its 541,516 inhabitants in 1900. It should be noted too that almost a century earlier fully 43% of the capital’s residents were also born outside the city and of these almost two-thirds were from neighboring districts. Assisted in part by its small geographical size, Mexico City was the strongest migratory magnet in the Mexican nation at the beginning of the twentieth-century, just as it had been one hundred years earlier for the viceroyalty of New Spain, and perhaps as long ago as half a millennium, as the center of the Mexica (Aztec) empire. This millennial-long process of urbanization at the core finally ended with the earthquake of 1985, after which the Federal District began to implode, expelling more migrants than it received.

Rapid urbanization in Mexico is a phenomenon of the twentieth-century. In 1900, only four cities, in addition to the Federal District, numbered as many as 50,000 inhabitants: Guadalajara (101,208), Puebla (93,521), Monterrey (62,266) and San Luis Potosí (61,019). For much of the nineteenth-century, to the extent that migration was important for urban growth, in-migrants were easily obtained from in-state. In the case of Guadalajara, for example, during the period 1875 to 1905 80% of the city’s migrants came from within the state according to data from city marriage registers. On the other hand, migrants were scarcely missed from the countryside because as late as 1900 72% of all Mexicans still lived in communities of fewer than 2,500 inhabitants (down from 92% a century earlier). Over three-fourths of the working class continued to depend upon agriculture for a livelihood while three-fourths of the middle class lived in towns and cities, which despite their small size offered migrants remunerative labor, schools, and a more open, egalitarian social structure.

Indian villages languished while haciendas, ranchos, and towns with predominantly Spanish speaking populations grew ever more rapidly. The declining number of Indian language speakers was due more to the spread of Spanish-based education (castillanización), changing identities and mestizaje than to lower reproduction rates. With the universal repudiation of racial distinctions after independence, the emergence of what the noted Mexican educator José Vasconcelos felicitously termed "the cosmic race" can be examined only through extant, but inexact statistics compiled at the beginning and end of the century. Thus, Afro-Mexicans, who numbered one-half million in 1810, more or less vanished, thoroughly intermingled and unidentifiable by 1895 if the official discourse is accepted at face value. Indians, according to the ill-defined 1793 count, numbered 2.5 million in central Mexico. A century later only 2.1 million Mexicans spoke Indian languages in the entire republic, a decline from at least fifty to a scant fifteen percent of the population.

If speakers of Indian languages had grown at the national average of one percent per annum over the century, they would have numbered 6.5 million in 1900. Since death rates in Indian regions were higher than the national average (Table 5), perhaps half of the four million loss could be attributed to higher mortality. The other half should be assigned to mestizaje and transformed identities, sometimes forced by gun toting, land-grabbers who found it convenient to make native villages disappear. The spread of public education, market economies, and liberal politics in the nineteenth-century proved as perilous to the survival of Native Americans, their cultures and communities, as virgin soil epidemics in the sixteenth.

In 1900 only six states reported one-fourth or more of total population as accounted for by native language speakers—Oaxaca, Chiapas, Guerrero, Campeche, Yucatán, and Puebla. By 1910, the group was halved to three. In absolute terms, 80% of all native language speakers in 1900 lived in seven states—50% in only three: Oaxaca, Puebla, and Yucatán (Figure 7). From 1900 to 1910, Indian language speakers increased their demographic weight in only three "states"—Chihuahua, Nayarit, and San Luis Potosí. By 1950 according to official figures, speakers of Indian languages numbered 2.4 million—of whom one-third were monolinguals. In recent decades a significant resurgence has taken place, reaching 5.2 million in 1980, due in part to the fact that acknowledging indigenous roots is increasingly a matter of pride rather than of shame as in earlier times. Nevertheless, by 1980 the fraction of native language speakers dropped below eight percent, barely half the figure at the beginning of the twentieth-century.

3.c. Life expectancy, disease, and the decline of epidemics

The slow acceleration of urbanization, migration and emigration in nineteenth-century Mexico was probably accompanied by modest gains in mortality. The nineteenth-century has long been written off as a loss in terms of mortality, a period in which life chances were poor and remained so over much of the century. Unfortunately, neither parish books nor civil registers are adequate for studying the course of life expectancy in Mexico until the twentieth-century. When sufficient information to hazard national estimates becomes available, life expectancy at birth is estimated at less than 30 years for both males and females. One might imagine that conditions could hardly have been worse at the beginning of the nineteenth-century. Nonetheless the thesis that mortality improved following independence is supported by four pieces of evidence: 1) growth rates implicit in national estimates of total population, 2) year-to-year fluctuations in mortality, 3) shifts in the mix of causes of death and 4) estimates of life expectancy for various times and places.

Since the first factor was already discussed above, consider the second here. The settling down of annual fluctuations in burials over the century suggests accelerating demographic growth rates. Great crises provoked in earlier times by typhus, smallpox and famine gradually tapered off. Mexico City serves as an example, although its series is confounded by a lack of true rates and the reality of demographic growth through migration (Figure 8). The last great crisis of colonial Mexico City was the typhus epidemic of 1813 when deaths more than tripled, rising from a greatly under-recorded annual average of five thousand to 17,021. When the biggest subsequent epidemic of typhus erupted eighty years later, the population of the capital had probably doubled, but typhus deaths rose to "only" 2,653, scarcely ten percent of total burials and a great public health victory by early modern standards. Although mortality remained high in Mexico City until the 1920s, annual fluctuations lost much of the volatility characteristic of the colonial regimen. Similar developments characterize burial series for Chiapas and Oaxaca in the south and Chihuahua in the north.

Third, consider changes in the causes of death over the century: the decline of mortality due to famine, smallpox, and other big killers of the colonial era. Hunger was certainly not absent from nineteenth-century Mexico (nor the twentieth either), but the frequency and intensity of famine seem to have lessened following the end of Spanish colonialism. Corn output probably doubled during the first half of the nineteenth-century while population increased by fifty percent. A Treatise on the Cultivation of Maize in Mexico written by the Mexican intellectual Don Luis de la Rosa and published in 1846 leaves no doubt that the most devastating hunger in Mexican history was that of 1786 (while the greatest in Mexican pre-history ravaged central Mexico in 1450). De La Rosa also expressed surprise and satisfaction at the great expanse of new lands brought under cultivation and sown in corn following the hunger of 1786. With independence, local, state and federal authorities sought to anticipate shortages, increase production by early harvests brought from the tropical lowlands, and insure that supplies were transported to areas of scarcity as needed—even during war years. Mexican intellectuals were pleased by this progress, but were displeased that most women continued to spend a large portion of their time at the metate because no means had been discovered for making fresh corn tortillas with the economies of baking wheat bread. Day-old tortillas acquired the texture of parchment, whereas day-old bread remained palatable.

In times of dearth, grain could be purchased from abroad, particularly from the United States, as in 1845 and 1909. In the Republic of Mexico, unlike in the colony of New Spain, during times of scarcity, there was rarely a doubling, much less quadrupling, of grain prices, a common phenomenon in colonial Mexico. Historians have long thought that in the decades prior to the Revolution of 1910 food production did not keep pace with accelerating population growth, but John Coatsworth’s meticulous critique of the evidence reveals this to be a myth. When reliable quantitative data become available, trends in per capita production of basic foodstuffs—corn, wheat, beans and even pulque (a native beverage, mildly intoxicating but rich in calories and vitamins)—were positive from 1877 to 1907, and increases were greatest when export agriculture was expanding most rapidly. The killing frosts of 1907 created shortages that were overcome by massive imports from the United States. Some historians cite this crisis as evidence of the failure of Mexican agrarian policies, which supposedly favored exports over domestic consumption. The fact that foodstuffs were readily obtained from the United States at reasonable prices prevented recurrence of another devastating famine such as that of 1786.

Mortality due to smallpox was greatly reduced following the introduction of vaccination in 1804. Thanks to royal patronage and enthusiastic endorsement by both clerical and secular elites, vaccination was extended quickly throughout the viceroyalty of New Spain. If smallpox was not eradicated in Mexico until 1951 (a decade prior to the last case in Great Britain), its virulence was substantially attenuated over the course of the nineteenth-century. In the epidemic which erupted in Mexico City in November 1829 for example, deaths increased "only" by three or four thousand (the surge was due in part to simultaneous outbreaks of measles, scarlet fever and dysentery) instead of ten or twelve thousand as was common in the previous century (see Figure 8). The monetary costs of stemming the tide of this epidemic, by hurriedly vaccinating the populace, was a fraction of the expense for care and hospitalization incurred during the epidemics of 1779 and 1797. Again in 1840 and 1844 mortality was greatly reduced by vaccination campaigns.

Historians looking back on the failure of Mexican authorities to eradicate smallpox by universal vaccination cite administrative chaos, incompetence, negligence, and lack of foresight as causes. Likewise, the transformation of smallpox into an endemic disease is interpreted as one of the great failures in Mexican public health. Not until 1917 was a national public health authority created. Then too, until 1917 Mexican doctors uniformly practiced the dangerous arm-to-arm method of vaccination, which demanded professional supervision (and fees) to reduce the risks of transmitting other communicable diseases such as syphilis. While a comprehensive history of smallpox will clear up many of the uncertainties about the demographic, social, and health significance of the disease, my reading of the evidence is that smallpox mortality declined sharply with the introduction of vaccination in 1804. Indeed the rapid growth of Mexico City in the late nineteenth-century was due in part to an unrelenting vaccination campaign, in which 20-50,000 children were treated annually, and smallpox deaths were reduced to a few hundred. When epidemic threatened, authorities in Mexico City rushed vials of vaccine (of human origin) to the hinterland to curtail the eruption. According to official statistics for the years 1894 to 1903 more than 3.5 million Mexicans were vaccinated against smallpox, yet 215,578 died from the disease in those same years. Thus at the end of the century smallpox accounted annually for three or four percent of all deaths. Measles and other big killers from colonial times were also tamed in the nineteenth-century, although the causes remain unclear.

On the other hand, war and cholera were new to Mexico in the nineteenth-century. The demographic costs of the former were substantial (and relatively unstudied), but the impact of cholera was relatively minor (and widely researched). Following conquest and during three centuries of colonial rule, there were few deaths from war, insurrection, or riot. The demographic catastrophe of the sixteenth-century was mainly due to disease and exploitation, not war. On the northern frontier, in the "provincias internas" north from Durango, fighting continued into the nineteenth-century and the destruction of native settlements was widespread, yet the demographic costs for the colony as a whole were minor. In the nineteenth-century, struggles for independence, civil wars and invasions—first by the United States in 1846 and then France in 1863—cost perhaps a half million lives—and another half million displacements as people permanently fled the war zones. The deadliest war of all was the decade of revolution and civil struggle that began in 1910. The assassination of President Francisco Madero in February 1913 and the ensuing confrontations of a score of regional bands inflicted massive destruction on the country.

Cholera found virgin soil in Mexico in 1833, following outbreaks in Western Europe and North America. The disease flared again in 1850 and 1882, yet Mexican authorities anticipating the epidemic’s appearance always sought the latest (at times, ineffective) methods from Europe to attempt to contain the spread and reduce mortality (Figure 8). Cholera provoked great horror in the populace, afflicting adults as well as children, with excruciatingly painful symptoms, but the overall demographic impact was relatively minor. In Mexico City the number of deaths increased ten thousand in 1833 and by somewhat fewer in 1850, but these epidemics failed to penetrate sparsely populated rural areas or much of the hinterland.

Life expectancy in nineteenth-century Mexico is difficult to estimate, but the few figures that we have point to significant improvements over the course of the century. A series for a parish with no war mortality during the nineteenth-century, Hidalgo del Parral in Chihuahua state, suggests an increase in life expectancy at birth from less than twenty years during the last decades of colonial rule to almost thirty years for the first decades of the republic and nearly forty years after mid-century, but falling below twenty again during the decade of revolution and the Spanish influenza epidemic of 1918-1919 (Table 5).

In rural Jalisco life expectancy at age five (e5) rose from 29 years in 1845-54 to 34 years in 1880 and 37 by 1900 (and 47 by 1950). The city of Guadalajara seems to have enjoyed an advantage of perhaps a year or two over its hinterland, while rural Oaxaca lagged behind Jalisco by four or five years. National estimates of life expectancy at birth for 1900 range between 25 and 30 years. One might imagine that the colonial record could not have been worse, but the life expectancy figures for Parral and the few annual burial series that we have for central Mexico convince me that during the last century there were substantial improvements in life chances, perhaps by as much as five or ten years. If true, this signaled a prolongation of the average lifespan by at least one-fifth and perhaps as much as one-half above colonial levels.

3.d. Fertility and marriageways

Birth rates for early modern Mexico are even more difficult to plot than death rates, because of the notorious shortcomings of parish and civil registration systems. In the Mixteca Alta of Oaxaca, a series developed by Cook and Borah reveals a century-long see-sawing between 45 births per thousand population in 1770 and 52 for the first decades of the republic. For 1900 demographers settle on 50 as a plausible crude birth rate for the nation as a whole. This corresponds to a total fertility rate of 6.8 children for women who survived to the end of their childbearing years. In other words, fertility was high in nineteenth-century Mexico, but trends and fluctuations remain obscure.

Fertility probably declined over the century, not because of any conscious effort to limit births but rather from, on the one hand, a delay in the age at marriage (including all "marriageways"—religious, civil and consensual), and, on the other, from a decrease in the proportion of women who entered unions of any sort. In late colonial Mexico, marriage was nearly universal. A church census for the Archbishopric of Mexico covering more than one million people reveals that in 1779 89% of females of marriageable age were married or widowed. A century and a half later, according to the 1930 census (the first enumeration to consider all forms of unions), the figure had fallen twenty points to 69% (compared with 73% in the U.S.A. in the same year). Concurrently, marriage age rose over the century. A study of Oaxaca’s northern districts documents a sustained increase in average female age at marriage from 16.2 years in 1700 to 18.9 years in 1905. In the Bajío the median age rose from 17.0 years in 1782-85 to 18.6 in the 1850s. In 1930 the national mean age at first union for females was 21.9 years and the proportion of women ultimately entering unions (by age 50) was 87%—surprisingly high in comparison to Europe or the United States, but low when compared to Mexico’s own historical record. (After 1930, the first figure would remain stable over the ensuing half century, while the proportion of females entering unions would actually rise to 93%, and the already high birth rate would surge by several points before plummeting in the 1970s.)

In the waning years of Spanish rule the position of women in the marriage market was undermined by a royal edict which decreed that only written pledges of marriage were legally binding. Unfortunately for women this restriction, issued in 1803, was carried over into republican civil codes. According to colonial custom among the popular classes, a spoken promise to marry by the male was followed by the female "losing her virginity to him" (from the male’s viewpoint, "[I] took her virginity"). After a reasonable interval, marriage usually followed. Church and civil courts stood ready to guarantee that a dallying suitor either fulfilled his promise or compensated the deflowered maiden. The court records show that notarized pledges were rarely obtained. After the edict of 1803, verbal seductions continued, but women were now stripped of the right to demand redress in the courts. In 1857, the institution of marriage was further weakened when civil marriage became the only form of legal wedlock in Mexico. Almost a century passed before the civil act became widely accepted.

4. Reflections

On November 25, 1910 Mexico’s long nineteenth-century abruptly ended with the outbreak of violent revolution that would ultimately destroy hundreds of thousands of lives and impel as many others to flee to the United States. Mexico’s high-pressure system in 1910, notwithstanding the demographic changes occurring over the previous century, was more akin to what had existed when Mexicans fought for independence from Spain—indeed closer to what the Mexica experienced before Christians invaded almost 400 years earlier—than to conditions of the late twentieth-century. Spanish colonization during the sixteenth-century as well as Liberal reforms in the nineteenth destabilized, but did not destroy, the fundamental population dynamics of agrarian societies of ancient Mexico. This would be accomplished only after 1930, by the greatest demographic transformation to occur in Mesoamerica since first settlement ten to seventy thousand years ago.

The millennial model of vital rates hypothesized in this essay is one of high but slowly easing pressure. Birth rates drifted downward from perhaps 60-70 per thousand population per annum in the fifteenth-century, to 55-65 in the eighteenth and 50-55 in the nineteenth (and 20-25 by the end of the twentieth). In normal years—of which there were few—death rates moved in tandem with births, but lagged a few points below. Beginning in the seventeenth-century the gap slowly widened, and by the late nineteenth spread to ten or fifteen points, as death rates continued their downward drift, finally bottoming out at less than seven deaths per thousand population by the year 2000. Birth rates declined too, although not as fast or as far. The decrease was due, at least prior to the twentieth-century fertility transition, to changing marriageways—not to birth control within marriage. With women marrying "later," although still relatively early compared to rural folk of Western Europe or colonial British America, and widows confined to a secondary marriage market, sexual activity languished for a measurable fraction of the female population. Birth rates declined accordingly, although they remained much higher than anything seen in Western Europe since the Middle Ages.

Malthusian crises are often perceived as the underlying causes of great Mesoamerican political cataclysms—conquest, independence wars, and revolution—over the past half millennium. Demographic determinism for even one of these upheavals is not convincing. Each was a piece of a general process that impacted peoples throughout the Americas. Population growth provided fodder for political catastrophe, but in each instance hemispheric forces were at work, so that sparsely populated regions, as well as densely settled ones, suffered the consequences of conquest, war and rebellion. What is astonishing is that once each catastrophe subsided, whether in Mesoamerica or the Southern Cone of South America, population expansion resumed, invariably attaining greater densities and greater social complexity which, in turn, established the foundation for further demographic growth without the menace of Malthusian collapse. Demographic pressure over the millennia was important and inter-related with environmental, political and even cultural change—but was rarely the sole determinant, just as it is unlikely to be in the next millennium.

Works Cited.

Aguirre-Beltrán, Gonzalo. La población negra de México. Estudio etnohistórico. México, DF: Fondo de Cultura Ecónomica, 1972 [1st ed. 1946].

Alaman, Lucas. Historia de Méjico. México, DF: Editorial Jus, 1942.

Alba, Francisco. "Population and the Crisis of the Socio-Political System: The Case of Prerevolutionary Mexico." in Eric Vilquin (ed.), Revolution et Population: Aspects démographiques des grandes révolutions politiques. Louvain-la-Neuve: Academia, 1990, pp. 203-218.

Anderson, Rodney. "Cambios sociales y económicos en el 6º cuartel de Guadalajara: 1842-1888." Encuentro El Colegio de Jalisco 1, no. 4 (jul-sep 1984): pp. 17-37.

Archivo del Ayuntamiento de la Ciudad de México (AACM), Estadística.

Archivo General de Indias (AGI). Fondo Varios.

Archivo General de la Nación (AGN). Ramo de Gobernación.

Archivo Histórico de la Secretaría de la Salubridad (AHSS). Fondo Salubridad Pública.

Archivo Parroquial de San José de Parral (APSJP), Libros de Defunciones.

Arriaga, Eduardo A. New Life Tables for Latin American Populations in the Nineteenth and Twentieth Centuries. Berkeley: University of California Institute of International Studies, 1968.

Arrom, Silvia M. "Changes in Mexican Family Law in the Nineteenth-century: The Civil Codes of 1870 and 1884." Journal of Family History 7, no. 3 (Fall 1985): pp. 305-317.

Arrom, Silvia M. The Women of Mexico City, 1790-1857. Stanford: Stanford University Press, 1985.

Azcarate, Miguel Maria del. Noticias Estadísticas que sobre los efectos de consumo introducidos en esta capital en el quinquenio de 1834 a 1838... Mexico, 1839.

Bauer, Arnold J. "Millers and Grinders: Technology and Household Economy in Meso-America." Agricultural History 64, no. 1 (Winter 1990): pp. 1-17.

Bennett, Herman Lee. Lovers, Family and Friends: The Formation of Afro-Mexico, 1580-1810. Unpub. PhD. thesis, Duke University, 1993.

Berdan, Frances F. and Patricia Anawalt (eds.). The Codex Mendoza. Berkeley: University of California Press, 1992.

Borah, Woodrow and Sherburne F. Cook. "La despoblación del México central en el siglo XVI." Historia Mexicana 12, no. 1 (jul-sep 1962): pp. 1-12.

Borah, Woodrow and Sherburne F. Cook. "Marriage and Legitimacy in Mexican Culture: Mexico and California." California Law Review 54, no. 2, (1966): 946-1008.

Borah, Woodrow and Sherburne F. Cook. The Aboriginal Population of Central Mexico on the Eve of the Spanish Conquest. Berkeley: University of California Press, 1963.

Borah, Woodrow and Sherburne F. Cook. The Indian Population of Central Mexico 1531-1610. Berkeley: University of California Press, 1960.

Brachet de Márquez, Viviane; Nettel, M. La población de los estados mexicanos en el siglo XIX, 1824-1895. México, DF: Instituto Nacional de Antropología e Historia, 1976.

Brading, David A. Haciendas and Ranchos in the Mexican Bajío: León, 1700-1860. Cambridge (England): Cambridge University Press, 1978.

Burkholder, Mark A. and Lyman L. Johnson. Colonial Latin America. Oxford: Oxford University Press, 1998 (3rd ed.).

Bustamante, Miguel E. "Aspectos históricos y epidemiológicos del hambre en México," in Enrique Florescano and Elsa Malvido, Ensayos sobre la historia de las epidemias en México. México, D.F.: Instituto Mexicano del Seguro Social, 1982, vol. 1, pp. 37-66.

Bustamante, Miguel E. "La situación epidemiológica de México en el siglo XIX," in Enrique Florescano and Elsa Malvido, Ensayos sobre la historia de las epidemias en México. México, D.F.: Instituto Mexicano del Seguro Social, 1982, vol. 2, pp. 425-476.

Bustamante, Miguel E. "La viruela en México, desde su orígen hasta su erradicación," in Enrique Florescano and Elsa Malvido, Ensayos sobre la historia de las epidemias en México. México, D.F.: Instituto Mexicano del Seguro Social, 1982, vol. 1, pp. 67-92.

Cabrera y Quintero, Cayetano. Escudo de Armas de Mexico: Celestial proteccion de esta nobilissima ciudad de la Nueva España, y de casi todo el nuevo Mundo, Maria Santissima, en su portentosa imagen del Mexicano Guadalupe, milagrosamente apparecida en el Palacio Arzobispal el año de 1531. y jurada su principal patrona el passado de 1737 en la angustia que ocasiono la pestilencia, que cebada con mayor rigor en los Indios, mitigo sus ardores al abrigo de tanta sembra. Mexico: 1746.

Calvo, Tomás. "Familles mexicaines au XVIIe siècle: une tentative de reconstitution." Annales de démographie historique 1984, 149-174.

Calvo, Tomás. Acatzingo: demografía de una parroquia mexicana. México, DF: INAH-SEP Colección Científica (Historia), 1973.

Calvo, Tomás. Guadalajara y su región en el siglo XVII: población y economía. Guadalajara: Centro de Estudios Mexicanos y Centroamericanos and H. Ayuntamiento de Guadalajara, 1992.

Calvo, Tomás. La Nueva Galicia en los siglos XVI y XVII. Guadalajara: El Colegio de Jalisco and Centro de Estudios Mexicanos y Centroamericanos, 1989.

Camposortega Cruz, Sergio. Análisis demográfico de la mortalidad en México, 1940-1980. México, DF: El Colegio de México, 1992.

Capello, Juan Francisco. Remedios contra la peste. Mexico 1737.

Carroll, Patrick J. "Los mexicanos negros, el mestizaje y los fundamentos olvidados de la ‘Raza Cósmica’: una perspectiva regional." Historia Mexicana 44, (ene-mar 1995): pp. 403-438.

Carroll, Patrick J. Blacks in Colonial Veracruz: Race, Ethnicity, and Regional Development. Austin: University of Texas Press, 1991.

Chavez Galindo, Ana Maria and Sandra Savenberg, "Le Centre du Mexique: De la Suburbanisation vers la Mégalopolisation." Population, 51, no. 3 (mai-juin 1996): pp. 756-766.

Ciudad de México. Manifiesto al publico que hace el ayuntamiento de 1840 acerca de la conducta que ha observado en los negocios municipales y del estado en que quedan los ramos de su cargo. México, D.F.: Impreso por Ignacio Cumplido, 1840.

Cline, Sarah L. The Book of Tributes. Early Sixteenth-Century Nahuatl Censuses from Morelos. Los Angeles: University of California Press, 1993.

Coale, Ansley and Paul Demeny. Regional Model Life Tables and Stable Populations. New York: Academic Press, 1983.

Coatsworth, John H. Los orígenes del atraso: Nueve ensayos de historia económica de México en los siglos XVIII y XIX. Mexico City: Alianza Editorial Mexicana, 1990.

Cohen, Mark Nathan. "Does Paleopathology Measure Community Health? A Rebuttal of ‘The Osteological Paradox’ and Its Implication for World History," in Richard R. Paine, Integrating Archaeological Demography: Multidisciplinary Approaches to Prehistoric Population. Carbondale IL: Southern Illinois University, 1997, pp. 242-260.

Cook, Sherburne F. "The Incidence and Significance of Disease Among the Aztecs and Related Tribes." Hispanic American Historical Reviews, 21, no. 3 (August 1946): pp. 320-335.

Cook, Sherburne F. "Migration as a Factor in the History of Mexican Population: Sample Data from West Central Mexico, 1793-1950," in Paul Deprez, ed., Population and Economics. Winnipeg: University of Manitoba Press, 1970, pp. 279-302.

Cook, Sherburne F. and Lesley Bird Simpson. The Population of Central Mexico in the Sixteenth-century. Berkeley: University of California Press, 1948.

Cook, Sherburne F. and Woodrow Borah. Essays in Population History: Mexico and the Caribbean. Berkeley: University of California Press, 1971-1979, 3 vols.

Cooper, D. B. Epidemic Disease in Mexico City, 1761-1813: An Administrative, Social and Medical Study. Austin: University of Texas Press, 1965.

Cope, Robert Douglas. The Limits of Racial Domination: Plebeian Society in Colonial Mexico City, 1660-1720. Madison: University of Wisconsin Press, 1994.

Cosío Villegas, Daniel. (ed.) Historia Moderna de México. México, DF: Editorial Hermes, 1955-72, 9 vols.

Davies, Keith A. "Tendencias demográficas urbanas durante el siglo XIX en México." Historia Mexicana, 21, no. 3 (ene-mar 1972): pp. 481-524.

De la Rosa, Luis. Memoria sobre el Cultivo del Maiz en Mexico. Mexico, 1846.

Denevan, William M., ed. The Native Population of the Americas in 1492. Madison: University of Wisconsin Press, 1976 [rev. ed. 1992].

Diamond, Jared. Guns, Germs, and Steel: The Fates of Human Societies. New York: W.W. Norton, 1997.

Doenges, Catherine E. "Patterns of Domestic Life in Colonial Mexico: Views from the Household." Latin American Population History Bulletin, no. 19 (Spring 1991), 14-21.

Erosa-Barbachano, Arturo. "La viruela, desde la Independencia (1821) hasta la erradicación," in Enrique Florescano and Elsa Malvido, Ensayos sobre la historia de las epidemias en México. México, D.F.: Instituto Mexicano del Seguro Social, 1982, vol. 2, pp. 545-550.

Faulhaber, Johanna. "La población de Tlatilco, México, caracterizada por sus entierros," in Homenaje a Juan Comas en su 65º aniversario. México, D.F.: Instituto Indigenista Interamericano, 1965, vol. 2, pp. 83-122.

Fernandez del Castillo, Francisco. Los viajes de Don Francisco Xavier de Balmis: notas para la historia de la expedición vacunal de España a América y Filipinas (1803-1806). México, D.F.: Galas de México, 1960.

Florescano, Enrique. Precios del maíz y crisis agrícolas en México (1708-1810): ensayo sobre el movimiento de los precios y sus consecuencias económicas y sociales. México, D.F.: El Colegio de México, 1969.

Genealogical Society of Utah (GSU), Salt Lake City, UT. Microfilm Collection of Parish Registers.

Gerhard, Peter. "La evolución del pueblo rural mexicano, 1519-1975." Historia Mexicana 24, no. 4 (abr-jun 1975): pp. 566-578.

Gerhard, Peter. Geografía Histórica de la Nueva España 1519-1821. México, DF: Universidad Nacional Autónoma de México, 1986 [1st English ed. 1972].

Gibson, Charles. The Aztecs Under Spanish Rule: A History of the Indians of the Valley of Mexico, 1519-1810. Stanford: Stanford University Press, 1964.

Gonzalez Navarro, Moisés. Los extranjeros en México y los mexicanos en el extranjero, 1821-1970. México, D.F.: Colegio de México and Centro de Estudios Históricos, 3 vols., 1993-1994.

Greer, Robert Gordon. The Demographic Impact of the Mexican Revolution. Austin: The University of Texas, 1966, unpub. thesis.

Hassig, Ross. "The Famine of One Rabbit: Ecological Causes and Social Consequences of a Pre-Columbian Calamity." Journal of Anthropological Research, 37 (1981): pp. 171-181.

Hayward, Michel H. A Demographic Study of Cholula, México, from the late Postclassic and the Colonial Period of 1642-1738. State University of Pennsylvania, 1986, unpub. PhD. thesis.

Henige, David. "Native American Population at Contact: Standards of Proof and Styles of Discourse in the Debate." Latin American Population History Bulletin, no. 22 (1992): pp. 2-23.

Hutchinson, Charles S. "The Asiatic Cholera Epidemic of 1833 in Mexico." Bulletin of the History of Medicine, 32, no. 1 (Jan-Feb 1958): pp. 1-23 and 152-163.

Humboldt, Alexander von. Essai politique sur le royaume de la Nouvelle-Espagne. Paris: F. Schoell, 1811.

Johansson, S. Ryan. "The Demographic History of the Native Peoples of North America: A Selective Bibliography." Yearbook of Physical Anthropology 25 (1982): pp. 133-152.

Johansson, S. Ryan, and S. Horowitz. "Estimating Mortality in Skeletal Populations: Influence of the Growth Rate on the Interpretation of Levels and Trends during the Transition to Agriculture." American Journal of Physical Anthropology 71 (1986): pp. 233-250.

Kanter, Deborah E. Hijos del Pueblo: Family, Community, and Gender in Rural Mexico, The Toluca Region, 1730-1830. Charlottesville: University of Virginia, 1993, Unpub. PhD. thesis.

Kicza, John E. "Mexican Demographic History of the Nineteenth-century: Evidence and Approaches." In James W. Wilkie and Stephen Haber (eds.), Statistical Abstract of Latin America. Los Angeles: University of California Press, UCLA Latin American Center Publications, 1981, vol. 21, pp. 592-609.

Klein, Herbert S. "The Demographic Structure of Mexico City in 1811." Journal of Urban History, 23, no. 1 (Nov. 1996): pp. 66-93.

Klein, Herbert S. "Familia y fertilidad en Amatenango, Chiapas (1785-1816)," in Elsa Malvido and Miguel Angel Cuenya, comp., Demografía Histórica de México: siglos XVI-XIX. México, DF: Universidad Autónoma Metropolitana and Instituto Mora, 1993, pp. 112-122.

Kubler, George. "Population Movements in Mexico, 1520-1600." Hispanic American Historical Review 22, no. 4 (Nov. 1942) pp. 606-643.

Kuznesof, Elizabeth Anne. "Gender Ideology, Race and Female Headed Households in Urban Mexico, 1750-1850," in Victor M. Uribe (ed.), State and Society in Spanish America During the ‘Age of Revolution’: New Research on Historical Continuities and Changes ca 1750s-1850s. Duke NC: Duke University Press, forthcoming.

Lerner, Victoria. "Consideraciones sobre la población de la Nueva España (1793-1810) Según Humboldt y Navarro y Noriega." Historia Mexicana, 17, no. 3 (ene-mar 1968): pp. 327-348.

Livi-Bacci, Massimo. A Concise History of World Population. Cambridge, Mass: Blackwell Publishers, 1993.

López de Velasco, Juan. Geografía y descripción universal de las Indias, Marcos Jimenez de la Espada (ed.). Madrid: Atlas, 1971.

MacNeish, R.S. "Social Implications of Changes in Population and Settlement Patterns of 12,000 Years of Prehistory in the Tehuacan Valley of Mexico" in P. Deprez (ed.), Population and Economics. Winnipeg: University of Manitoba Press, 1970, pp. 215-250.

MacNeish, R.S. The Prehistory of the Tehuacan Valley. Austin: University of Texas Press, 1967.

Maldonado, Celia L. Estadísticas vitales de la ciudad de México. Siglo XIX. México, DF: Instituto Nacional de Antropología e Historia, 1976.

Malvido, Elsa. "Factores de despoblación y de reposición de la población de Cholula (1641-1810)." Historia Mexicana 23, no. 1 (jul-sep 1973): pp. 52-110.

Malvido, Elsa and Carlos Viesca. "La epidemia de cocoliztli de 1576." Historia, no. 11 (1985): pp. 27-33.

Márquez Morfín, Lourdes. La desigualdad ante la muerte en la Ciudad de México: el tifo y el cólera (1813-1833). México DF: Siglo XXI, 1994.

Márquez Morfín, Lourdes. "La evolución cuantitativa de la población novohispana: siglos XVI, XVII y XVIII," in El Poblamiento de México. México, DF: Consejo Nacional de la Población, 1993, vol. 2, pp. 36-63.

Marquéz Morfín, Lourdes, M.E. Peraza, J. Gamboa, and T. Miranda. Playa del Carmen: Una población de la costa oriental en el postclásico (un estudio osteológico). México, DF: Instituto Nacional de Antropología e Historia, Colección científica 119, 1982.

McCaa, Robert. "Calidad, Clase, and Endogamy in Colonial Mexico: The Case of Parral, 1788-1790." Hispanic American Historical Review, 64, no. 3 (August 1984): pp. 477-502.

McCaa, Robert. "Descenso de la fecundidad y modos de uniones matrimoniales en México: casos de Chihuahua y Puebla." in Actas de la VII Jornada Nacional de Historia Regional de Chile 1996. Santiago: Universidad de Chile, Departamento de Ciencias Históricas, pp. 59-78.

McCaa, Robert. "¿Fue el siglo XVI una catástrofe demográfica para México? Una respuesta basada en la demografía histórica no cuantitativa." Cuadernos de Historia, 15 (dic 1995): pp. 123-136.

McCaa, Robert. "La viuda viva del México borbónico: Sus voces, variedades y vejaciones." in Familias novohispanas Siglos XVI al XIX, Pilar Gonzalbo Aizpuru (ed.). México, DF: El Colegio de México, 1991, pp. 299-324.

McCaa, Robert. "Marriageways in Mexico and Spain, 1500-1900." Continuity and Change 9, no. 1 (1994): pp. 11-44

McCaa, Robert. "Matrimonio infantil, cemithualtin (familias complejas), y el antiguo pueblo nahua." Historia Mexicana 46, no. 1 (jul-sep. 1996): pp. 3-70.

McCaa, Robert. "Spanish and Nahuatl Views on Smallpox and Demographic Catastrophe in the Conquest of Mexico." Journal of Interdisciplinary History 25, no. 3 (Winter 1995): pp. 397-431.

McCaa, Robert. "The Peopling of Nineteenth-Century Mexico: Critical Scrutiny of a Censured Century," in James W. Wilkie, Carlos Alberto Contreras, and Christof Anders Weber (eds.), Statistical Abstract of Latin America. Los Angeles CA: University of California at Los Angeles Latin American Studies Center, 1993, vol. 30, part 1, pp. 603-633.

McCaa, Robert. "Women’s Position, Family and Fertility Decline in Parral (Mexico) 1777-1930." Annales de Démographie Historique (1989): pp. 233-243.

McCaa, Robert. and Márquez Morfín, Lourdes. "Paleodemography, Nutrition and Health in Ancient Mexico," unpublished paper presented at the Social Science History Association annual meeting, Chicago IL, November 16, 1995.

McClung, E. and M.C. Serra Puche. "La revolución agrícola y las primeras poblaciones aldeanas," in El Poblamiento de México. México, DF: Consejo Nacional de la Población, 1993, vol. 1, pp. 138-163.

México. Departamento de la Estadística Nacional (DEN). Censo de población. 15 de mayo de 1930. México, D.F.: Talleres gráficos de la nación, 1932.

México. Dirección General de Estadística (DGE). Resumen general del censo de de la República Mexicana verificado el 28 de octubre de 1900. México, DF: Secretaría de Fomento, 1905.

México. Dirección General de Estadística (DGE). Boletín demográfico de la República Mexicana, 1903. México, D.F.: Oficina Tip. de la Secretaría de Fomento, 1908.

Mexico. Instituto Nacional de Estádistica, Geografía e Informática (INEGI). Población y vivienda. Muestra estadística del censo de 1990. Aguascalientes: INEGI, 1994.

México. Instituto Nacional de Estadística, Geografía e Informática (INEGI). Estadísticas Históricas Mexicanas. México, DF: INEGI, 1986, 2v.

México. Instituto Nacional de Estadística, Geografía e Informática (INEGI). Primer censo de población de la Nueva España 1790. Censo de Revillagigedo, un censo condenado. México, DF: INEGI, 1977.

México. Secretaría de Estado. Memoria de la Secretaría de Estado y del Despacho de Relaciones Exteriores e Interiores, correspondiente al año de 1831. México, D.F.: 1831.

Mier y Terán, Marta. Evolution de la population mexicaine à partir des données des recensements, 1895-1970. Montréal: Université de Montréal, Département de Démographie, 1982, Unpub. thesis.

Mier y Terán, Marta. "Evolución demográfica de México en el siglo XX," in Sérgio Odilon Nadalin, Maria Luiza Marcílio and Altiva Pillati Balhana (eds.), História e População: Estudos sobre a América Latina. São Paulo: Fundação Sistema Estadual de Análise de Dados, 1990, pp. 81-87.

Moreno Toscano, Alejandra. "Cambios en los patrones de urbanización en México, 1810-1910." Historia Mexicana, 22, no. 2 (1972): pp. 160-187.

Morin, Claude. "Démographie et différences ethniques en Amérique latine coloniale." Annales de démographie historique (1977): pp. 301-312.

Morin, Claude. "Des terres sans hommes aux hommes sans terres: Les paramètres agraires de l’évolution démographique dans l’Indoamérique (Mexique-Pérou)," in Antoinette Fauve-Chamoux (ed.), Evolution agraire et croissance démographique, Liège: Ordina Editions, 1987, pp. 75-87.

Morin, Claude. "Los libros parroquiales como fuente para la historia demográfica y social novohispana." Historia Mexicana 21, no. 3 (ene-mar 1972): pp. 389-418.

Morin, Claude. Michoacán en la Nueva España del siglo XVIII: Crecimiento y desigualdad en una economía colonial. México, DF: Fondo de Cultura Económica, 1979.

Morin, Claude. Santa Inés Zacatelco (1646-1812): Contribución a la demografía histórica del México colonial. México, DF: Instituto Nacional de Antropología e Historia, 1973.

Motolinía, Fray Toribio de Benavente. Memoriales: libro de oro (MS JGI 31), Nancy Joe Dyer (ed.). México, D.F.: El Colegio de México and Centro de Estudios Linguísticos y Literarios, 1996.

Navarro y Noriega, Fernando. Memoria sobre la población del reino de Nueva España. México: en la Oficina de D. Juan Bautista de Arizpe, 1820.

Oliver Sánchez, Lilia V. Un verano mortal. Análisis demográfico y social de una epidemia de cólera, Guadalajara 1833. Guadalajara, Jalisco: Unidad Editorial del Gobierno del Estado, 1986.

Ortiz de Montellano, B.R. Aztec Medicine, Health and Nutrition. New Brunswick, NJ: Rutgers University Press, 1990.

Othón de Mendizábal, Miguel. "Demografía mexicana. Epoca colonial 1519-1810. Demografía colonial del siglo XVI. 1519-1599," in Obras Completas, Mexico, DF: Talleres Gráficos de la Nación, 1946 [first published 1939], vol. 3, pp. 309-338.

Ouweneel, Arij. "Growth, Stagnation, and Migration: An Explorative Analysis of the Tributario series of Anáhuac (1720-1800)." Hispanic American Historical Review 71, no. 3(Aug 1991), 531-577.

Ouweneel, Arij. Shadows over Anáhuac: An Ecological Interpretation of Crisis and Development in Central Mexico 1730-1800. Albuquerque NM: University of New Mexico Press, 1996.

Pescador, Juan J. De Bautizados a fieles difuntos. México, DF: El Colegio de México, 1992.

Piñeyro, Salvador. Memoria del las Chiapas. San Cristoval, Chiapas, 1830.

Prem, Hanns J. "Disease Outbreaks in Central Mexico during the Sixteenth-Century," in Noble David Cook and W. G. Lovell (eds.), "Secret Judgments of God." Old World Disease in Colonial Spanish America. Norman, OK: University of Oklahoma Press, 1991, pp. 20-48.

Quilodrán Salgado, Julieta. Le Mariage au Mexique: évolution nationale et typologie régionale. Louvain-la-Neuve: Université Catholique de Louvain, 1996, unpub. thesis.

Rabell Romero, Cecilia A. La población novohispana a la luz de los registros parroquiales. México, DF: Universidad Autónoma Nacional de México, 1990.

Rabell Romero, Cecilia. A.; Necochea, N. "La mortalidad adulta en una parroquia rural novohispana durante el siglo XVIII." Historia Mexicana 36, no. 3 (ene-mar 1987): pp. 405-442.

Ramos Escandón, Carmen. "Mujeres trabajadoras en el México porfiriano: Genero e ideología del trabajo femenino 1876-1911." Revista Europea de Estudios Latinoamericanos y del Caribe 48 (Jun 1990): pp. 27-45.

Reff, Daniel T. Disease, Depopulation, and Culture Change in Northwestern New Spain, 1518-1764. Salt Lake City UT: University of Utah Press, 1991.

Reher, David H. "¿Malthus de nuevo? Población y economía en México durante el siglo XVIII." Historia Mexicana 41, no. 4 (abr-jun 1992): pp. 615-664.

Robinson, David J. Research Inventory of the Mexican Collection of Colonial Parish Registers. Salt Lake City: University of Utah Press, 1980.

Rosenblat, Angel. "The Population of Hispaniola at the Time of Columbus," in The Native Population of the Americas in 1492, edited by William Denevan. Madison: University of Wisconsin Press, 1976 [rev. ed. 1992], pp. 43-66.

Rosenblat, Angel. La población indígena y el mestizaje en América. Buenos Aires: Editorial Nova, 1954, 2 vols.

Sahagún, Fray Bernardino de. Florentine Codex General History of the Things of New Spain, trans. Arthur J. O. Anderson and Charles E. Dibble. Salt Lake City, UT and Santa Fe, NM: University of Utah Press and School of American Research, Santa Fe, 13 vols., 1950-82.

Sánchez-Albornoz, Nicolás. La población de América Latina: Desde los tiempos precolombinos al año 2025. Madrid: Alianza Editorial, 1994.

Sanders, William T., Jeffrey R. Parsons, and R.S. Santley. The Basin of Mexico: Ecological Processes in the Evolution of a Civilization. New York: Academic Press, 1979.

Sanders, William T. "Ecological Adaptation in the Basin of Mexico 23,000 B.C. to the Present," in Supplement to the Handbook of Middle American Indians, Ronald Spores (ed.). Austin: University of Texas Press, 1986, vol. 4, pp. 147-97.

Sanders, William T. "The Population of the Central Mexican Symbiotic Region, the Basin of Mexico, and the Teotihuacan Valley in the Sixteenth-century," in William Denevan (ed.), The Native Population of the Americas in 1492. Madison: University of Wisconsin Press, 1976 [rev. ed. 1992], pp. 85-150.

Saul, F. The Human Skeletal Remains of Altar de Sacrificios: an Osteobiographic Analysis. Cambridge, Mass: Papers of the Peabody Museum, 1972, vol. 63, no. 2.

Serrano Sánchez, C. "Orígen del hombre americano y perfil biológico de la población prehispánica de México," in El Poblamiento de México. México, DF: Consejo Nacional de la Población, 1993, vol. 1, pp. 96-115.

Stern, Steven J. The Secret History of Gender: Women, Men, and Power in Late Colonial Mexico. Chapel Hill NC: University of North Carolina Press, 1995.

Storey, Rebecca. Life and Death in the Ancient City of Teotihuacan. A Modern Paleodemographic Synthesis. Tuscaloosa, AL: University of Alabama Press, 1992.

Sugiura Yamamoto, Y. "El ocaso de las ciudades y los movimientos poblacionales en el Altiplano Central," in El Poblamiento de México. México, DF: Consejo Nacional de la Población, 1993, vol. 1, pp. 190-213.

Swann, Michael M. Migrants in the Mexican North: Mobility, Economy, and Society in a Colonial World. Boulder CO: Westview Press, 1989.

Thompson, Angela. "To Save the Children: Smallpox Inoculation, Vaccination, and Public Health in Guanajuato, Mexico, 1797-1840." The Americas 49, no. 4 (April 1993): pp. 431-455.

Tribunal del Consulado. "Noticias de Nueva España en 1805." Boletín de la Sociedad Mexicana de Geografía y Estadística 2 (1850): pp. 5-13.

United States Bureau of the Census. Fifteenth Census of the United States, 1930: Population. General Reports, Statistics by Subjects. Washington, 1933, vol. 2.

Urías Hermosillo, Miguel and San Juan Victoria, Carmen "Población y desarrollo en el México del siglo XIX." Investigación Económica 162 (oct-dic 1982): pp. 129-177.

Van Young, Eric. Hacienda and Market in Eighteenth-Century Mexico. The Rural Economy of the Guadalajara Region, 1675-1820. Berkeley: University of California Press, 1981.

Viesca T., Carlos. "Hambruna y epidemia en Anáhuac (1540-1454) en la época de Moctezuma Ilhuicamina," in Enrique Florescano and Elsa Malvido (eds.), Ensayos sobre la historia de las epidemias en México. México, D.F.: Instituto Mexicano del Seguro Social, 1982, vol. 1, pp. 157-168.

Viesca T., Carlos. "Epidemiologia entre los mexicas," in Fernando Martínez Cortés (ed.), Historia General de la medicina en Mexico. México, D.F.: Universidad Nacional Autónoma de México, 1984, vol. 1, pp. 171-188.

Villaseñor y Sánchez, José Antonio. Theatro Americano: Descripción general de los reynos y provincianos de la Nueva España y sus jurisdicciones. México, DF: Editorial Trillas, 1992.

Wallace, Douglas C. "Mitochondrial DNA in Aging and Disease." Scientific American 277, no. 2 (August 1997): pp. 40-47.

Walter, John and Roger Schofield. "Famine, Disease and Crisis Mortality in Early Modern Society," in Walter and Schofield (eds.) Famine, Disease and the Social Order in Early Modern Society. Cambridge, Engl.: Cambridge University Press, 1989, pp. 1-74.

Whitmore, Thomas M. Disease and Death in Early Colonial Mexico: Simulating Amerindian Depopulation. Boulder, CO: Westview Press, 1992.

Wrigley, E. A. and Roger S. Schofield. The Population History of England, 1541-1871: A Reconstruction. London: Edward Arnold, 1981.

Zambardino, Rudolph A. "Mexico’s Population in the Sixteenth-century: Demographic Anomaly or Mathematical Illusion." Journal of Interdisciplinary History 11, no. 1 (Summer 1980): pp. 1-27.

Zavala de Cosío, María Eugenia. Cambios de fecundidad en México y políticas de población. México DF: El Colegio de México y Fondo de Cultura Económica, 1992.

Table 1. Mean age at death
at selected ages:

Mesoamerican precontact populations and others
at various economic-technological levels
mean age at death in years

Birth (age 0)
Age 15
Age 50
Tlatilco, 2930-3250 BPa323754
La Ventanilla, 350-950a364054
Cholula, 850-1560a293554
Copán rural, 700-1000a254460
Copán urban, 700-1000a364157
Teotihuacán, early classicb2442 58
Teotihuacán, late classicb163466
Teotihuacán, 1580-1620b132852
Cholula, 1325-1520c253451
North American Indiansd223555
Primitive agriculturalistsd264568
Model life table, South level 1e204965
Mexico, 1939-41f405870
Sources: aHealth and Nutrition in the Western Hemisphere Database, October 12, 1995;
bStorey Life and Death, pp. 184-5;
cHayward, Demographic study, pp. 221-2 (Table 7.5);
dJohansson, "Demographic History," p. 136;
eCoale and Demeny, Regional Model Life Tables, p. 384;
fCamposortega Cruz, Análisis, p. 321.
Note: Calculations are based on conventional paleodemographic assumptions: skeletal remains are a random sample of deaths for the population studied; the population is closed (migration is nil) and static (crude birth and death rates are equal; therefore the rate of population change is zero). Under these conditions, the mean age at death from age x is equivalent to life expectancy at that age.
Recently paleodemographers have begun to accept the fact that fertility, not mortality, is the principal determinant of the age structure of a population, and hence, of the age structure of deaths (Johansson and Horowitz 1986). The practical significance of this change in thinking is nil. As long as the assumption of a static system is retained, whether one first computes fertility and then derives mortality or the reverse has no practical effect. Both approaches yield the same answer. Problems of bias remain (see also note 19).

Table 2. Demographic Disaster in Mexico

Authoritative estimates of Total Population
and Implied Rates of Decrease

percent decrease
Cook and Simpson10.52.1-3.071-80
Cook and Borah18-301.478-95
SandersCentral Mexican
Symbiotic Region
WhitmoreValley of Mexico1.3-2.70.1-0.469-96
Kubler128 towns0.20.150
Sources: Rosenblat, Población indígena, vol. 1, pp. 57-122.
Aguirre-Beltrán, Población negra, pp. 200-1, 212.
Zambardino, "Mexico’s Population," pp. 21-2.
Mendizábal, "Demografía," vol. 3, p. 320.
Cook and Simpson, Population, pp. 38, 43, 45.
Cook and Borah, Aboriginal Population, p. 88.
Cook and Borah, Indian Population, pp. 46-7 (as corrected).
Sanders, "Central Mexican Symbiotic Region," p. 120; "Ecological Adaptation," p. 194.
Whitmore, Disease, p. 154.
Gibson, Aztecs, pp. 137-138.
Kubler, "Population Movements," p. 621.
Note: The nadir of the demographic disaster is usually placed in the seventeenth-century. I chose 1595 for an end-point, not because I think this to be the nadir of the native population, but to be able to interpolate, rather than extrapolate, comparable figures for the largest number of authors. Nevertheless, Sanders’ figure for the Valley of Mexico is extrapolated from 1568.

Table 3. Life expectancy at age 15 (e15)
colonial Mexico

place, periode15(years)author, method, data
Mexico, 1939-4143Camposortega, deaths, census
Model West, level 1 (e0=20)31Coale and Demeny, stable populations
Morelos, 1530s<30McCaa, orphanhood, children
Teotihuacán, 1580-162013Storey, mean age at death, skeletal remains
Cholula, 1642-9029Hayward, age at burial
San Luis de la Paz, 1745-9433Rabell, mean age at death, burial
... males35 (level 5)
... females31 (level 1)
Oaxaca, (1700-)177733Cook and Borah, population age structure, census
Parral, 1808<30McCaa, orphanhood, newlyweds
Sources: Camposortega Cruz, Análisis, p. 321.
Coale and Demeny, Regional Model Life Tables, p. 384.
Cook and Borah, Essays, vol. 1, pp. 201-299.
Hayward, "Demographic Study," pp. 221-2.
McCaa, "Peopling," pp. 617-8; "Matrimonio infantil," pp. 32-34.
Rabell, Población novohispana, p. 32.
Storey, Life and Death, pp. 184-5.
Example: In 1939-41, Mexicans aged 15 years could expect to live 43 additional years on average, to age 58, given the mortality experience of that period.

Table 4 Two series of population estimates
Mexico, 1790-1910

yearnew series1conventional seriesauthor
18105.66.1Navarro y Noriega3
18547.9Orozco y Berra3
189512.6I Censo3
190013.613.6II Censo3
191015.215.2III Censo3
Sources: 1McCaa, "Peopling," pp. 604-608 and as revised in this essay.
2Humboldt, Essai politique, vol. 1. pp. 325, 341;
3México INEGI, Estadísticas Históricas, Table 1.2.

Table 5 Life Expectancy in Mexico
both sexes combined

at age 5 (e5)180018301880190019301950
Oaxaca (rural) ....2933..40
Jalisco (rural) ....3437..47
Republic of Mexico......39..55
at age 0 (e0)1808182818781930
Sources: Cook and Borah, Essays, vol. 2, p. 398 figure 7.7;
Mier y Terán, Evolution, vol. 2, p. 196;
McCaa, "Peopling," pp. 617-619.