The Revillagigedo enumeration, 1790-1794: the first general count of New Spain
Before 1794 the archives of the viceroyalty of Mexico held no enumeration — the 1742 operation had estimated families, not persons. The count ordered by the second Count of Revillagigedo returned 3,865,529 souls in the jurisdictions finished by 1793 and about 4,483,529 with the three unfinished ones estimated. It was also an undercount, for a reason its own critic states: households shrank their lists for fear of a new tax.
Why this wins its question: English-language writing dates modern census-taking to the United States in 1790 and Britain in 1801 and almost never records that in the same years a general enumeration of persons was carried through most of a territory holding some four and a half million people, or that its results were printed and criticised within fifteen years. This object gives the table by intendancy as Humboldt published it from the viceroy's own copy, names the three jurisdictions where the count was never finished and the estimates substituted for them, and — instead of quoting the round 4,483,529 as a fact — puts on the record the fiscal fear that made households shrink their lists, the capital's 112,926 against an indicated 135,000, the rival copy giving 638,771 for Puebla, and the ten years in which intendants recopied a decade-old table as if nobody had been born.
Claims
Every assertion below is bound to registered sources and carries its own confidence. Weight them; do not treat the page as uniformly authoritative.
There was no earlier enumeration to compare it with. Humboldt, who worked in the viceroyalty in 1803 and 1804, writes that "the archives of the viceroyalty of Mexico contain no enumeration before 1794, when the Count de Revillagigedo, one of the wisest and most active administrators, had resolution enough to undertake it"; that in the operations ordered by the viceroy Count de Fuenclara in 1742 "the number of families only was estimated"; and that what Villaseñor preserved of that operation "is both incomplete and inaccurate" (vol. I, p. 96).
Villaseñor's own printed page confirms the unit and the categories. Describing Pátzcuaro in the second volume of the Theatro americano, the survey drawn up on the royal order of 19 July 1741 and printed in 1748, he writes: "Consta su Vecindario de mas de quinientas familias de Españoles, Mestizos, y Mulatos, y la Republica de los Indios con su Gobernador, y Alcaldes se compone de dos mil familias del Idioma Tarasco" (parte segunda, p. 14). Families rather than persons, and the two repúblicas counted apart — which is what a count of souls in 1790 replaced.
The result was published as a table of the intendancies and governments "in which the enumeration was completed in 1793", from the notices transmitted by the intendants and governors to the viceroy before 12 May 1794. Its entries run Mexico 1,162,886 (capital 112,926), Puebla 560,443, Tlaxcala 59,177, Oaxaca 411,336, Valladolid 289,314, Guanajuato 397,924, San Luis Potosí 242,280, Zacatecas 118,027, Durango 122,866, Sonora 93,396, New Mexico 30,953, the two Californias 12,666 and Yucatán 358,461, for a total deduced from the enumeration of 1793 of 3,865,529 (vol. I, p. 97).
Three jurisdictions were never finished, and the round total everyone quotes is partly an estimate rather than a count. The enumeration "was not completed in the two intendancies of Guadalaxara and Vera Cruz, and in the small province of Cohahuila"; in a report to the king Revillagigedo put Guadalajara at 485,000, Veracruz at 120,000 and Coahuila at 13,000, and it is the addition of those 618,000 estimated souls to the 3,865,529 counted that yields the "approximative result of the enumeration in 1793" of 4,483,529 inhabitants (vol. I, pp. 96-97).
The count was an undercount, and its critic gives the mechanism rather than a guess. "In the new continent, as well as in the old, every enumeration is considered by the people as a sinister presage of some financial operation. In the fear of an augmentation of imposts, every head of a family endeavoured to diminish the number of individuals of his house, of which he was to furnish a list." The test offered is the capital: consumption tables, births and burials indicated a population above 135,000, and the table printed by order of the viceroy in 1790 exhibits only 112,926; "in smaller cities, easier to be controuled, the error was still more considerable" (vol. I, p. 98).
Its own transmission was unstable, which is the strongest reason to treat any single figure with care. Humboldt says he published the state "from a copy preserved in the archives of the viceroy" and adds that he "observed that other copies in circulation in the country contain different numbers; for example, 638,771 souls for the intendancy of la Puebla, including the ancient republic of Tlascala" — a variant that exceeds this table's Puebla and Tlaxcala entries taken together (vol. I, p. 97).
The instrument was not maintained. "The viceroys who succeeded to the Count de Revillagigedo have never renewed the enumeration; and since that time, the government has paid very little attention to statistical researches. Several memoirs drawn up by intendants on the actual state of the country confided to their care contain exactly the same numbers as the table of 1793, as if the population could have remained the same for ten years." Correcting for those who withdrew themselves, a sixth or a seventh part was supposed to require adding, and the population of all New Spain was accordingly estimated at 5,200,000 souls (vol. I, p. 99).
What this evidence is, and what it cannot settle. Everything above is a Prussian traveller's report, printed at London in 1811, of a manuscript state he was shown in the viceroy's archives; the padrón itself was not opened for this object, and the returns of the parishes and subdelegaciones beneath the table are not in view. Humboldt is also an open advocate of the colonial scientific establishment and writes with his own argument in hand — the same page that reports the unrenewed enumeration ends in a peroration on fanaticism and the avarice of the first conquerors. Nothing here shows how the count was executed in the villages, whom the caste categories absorbed or excluded, or how the tributary rolls of the repúblicas de indios were reconciled with it.