A census is an official count or survey of a population that typically records various details about the individuals enumerated. Thus, the numerous estimates of the population of Texas in its early years, such as the one in Henderson Yoakum's History of Texas (1855), were not censuses. Also, given that most Texas Indians roamed widely across the region and could not be formally counted in any census, their numbers are available only in estimates by contemporaries and calculations by ethnic demographers. For estimates of the numbers of people in early Texas, including those for the region’s American Indian residents (see POPULATION AND DEMOGRAPHIC CHANGE).
The first attempt at a true census of Texas came in 1777 in response to a Royal Order requiring periodic census reports for all the people of Spain’s American dominions. Census takers in 1777 reported a total population of 3,103 in Texas divided as follows: 2,060 in Bexar (1,351 persons listed as civilian or military and 709 as residents of missions), 696 in La Bahia (515 military or civilian and 181mission residents), and 347 in Nacogdoches (with no indication of occupation). During the twelve years following the census of 1777, Spanish officials ordered twelve additional general and local population counts in Texas. However, these censuses generally were poorly done, and only partial reports are available from them. Then, in 1791, Viceroy Conde de Revillagigedo ordered a new general census that came with detailed instructions and a printed model for use by the enumerators. This census, which was completed in 1793, reported data on population characteristics such as age, sex, place of birth, race, occupation, and marital status. There is an excellent summary and analysis of the censuses taken in 1777 through 1793 by Alicia V. Tjarks in the Southwestern Historical Quarterly, Volume LXXVII, January 1974.
During the years from 1793 through 1834 – as Texas evolved from a province in the declining Spanish empire to a state (in combination with Coahuila) in the independent Mexican nation – local authorities attempted to conduct censuses on an annual basis. The results are found in microfilm copies of the Bexar Archives collection in the Briscoe Center for American History at the University of Texas at Austin. A published compilation of the results of local censuses in this period is a 1959 book by Marian D. Mullins entitled The First Census of Texas, 1829‑1836: To which are added Texas Citizenship Lists, 1821‑1845. The title is misleading in that it ignores earlier Spanish censuses, but the work is a useful presentation of censuses taken in Stephen F. Austin's Colony, Nacogdoches, San Augustine, Bevil’s Settlement, and the Sabine area. Another compilation of local census results was published by Gifford White in 1983 under the title, 1830 Citizens of Texas: A Census of 6,500 Pre‑Revolutionary Texians. It contains the Austin Colony 1825 "Register of Families," and 1830 censuses of San Antonio and Nacogdoches.
The Republic of Texas did not conduct a census during its ten years from 1836 to 1846. There is a 1966 compilation by Gifford White entitled The 1840 Census of the Republic of Texas, which is a reproduction of the tax rolls of twenty-six of Texas's thirty-two counties in 1840. Tax rolls have much useful statistical information, but they are not population counts.
Soon after statehood, an act of the legislature resulted in an 1847 state census that was published in the appendix to Laws Passed by the Second Legislature of the Second Legislature of the State of Texas. This census had county-by-county reports on the White male population divided into age groups (under 18, over 18 and under 45, over 45), the White female population (not divided into age groups), the total population of enslaved people, and total “free colored persons.” When historian William R. Hogan edited the returns nearly a century later for publication in the Southwestern Historical Quarterly, he pointed out that although they were “incomplete and incorrect in some spots, the same is true of most census data.”
As Antebellum Texas drew to a close, a "Census of the State of Texas for 1858" appeared in the Texas Almanac for 1859. Based on returns from tax assessor/collectors, this census reported for most counties (a few did not send reports) the White population grouped according to gender and age classifications, the total enslaved population, and the total number of “free colored persons.” It included a count of “qualified electors,” children at school out of state, and handicapped persons such as the blind and deaf. There was even a partial census of agriculture that reported counts of acres of land planted in corn, wheat, cotton, and sugar, as well as the total number of acres under cultivation.
The first United States Census of Texas came in 1850, a fortunate timing because the Federal government greatly expanded its decennial census after 1840. Whereas enumerators who took the Census of 1840 had to complete only one printed schedule, in 1850 they carried six schedules as they began on June 1 to travel from one rural home to the next and to make their way through the villages and towns of Texas. There were separate schedules for Free Inhabitants, Slave Inhabitants, Mortality (a listing of persons who had died during the previous year), Productions of Agriculture, Productions of Manufactures, and Social Statistics. The first schedule included information such as name, age, gender, marital status, occupation, and value of real property owned when relevant for every individual in every household visited. The second schedule listed every slaveholder by name and recorded each enslaved person by age, gender, and color, but unfortunately not by name. The fourth schedule listed every farm by owner or operator and gave the acreage of the farm along with its livestock and crop production. Thus, the data on these schedules, when combined, gave a detailed “snapshot” of the people and economy of the state.
The manuscript copies of schedules one and two, called the “population schedules,” eventually were preserved in the National Archives, and the “non-population schedules” such as agriculture and manufactures were left to the care of individual states. Compilations of the data in aggregate form were published as soon as possible by the Federal Government. Texas data for 1850 appeared on pages 308-319 in Statistical View of the United States … Being a Compendium of the Seventh Census, which was compiled by James D. B. DeBow and published in Washington in 1854.
The Census of 1860 relied on essentially the same six schedules used in 1850. One important addition was a column on the first schedule for the value of personal property that, when combined with the value of real property, created a measure of total wealth for each individual. Slaves were considered personal property and usually comprised a large part of the wealth of the richest Texans at the close of the antebellum years. Compilations that included the 1860 data from Texas as well as all other states were published by the U. S. Bureau of the Census in four volumes in 1864.
The Census of 1870 marked the first population count after the Civil War and emancipation and significantly meant the elimination of Schedule 2 – Slave Inhabitants. Other than that, however, the schedules were the same as in 1860. Taken soon after the upheaval of war and during the controversies of Reconstruction, the 1870 census was notably incomplete – and often inaccurate – for Texas. The compiled returns from all states were published by the Bureau of the Census in three volumes.
Prior to the Census of 1880 major changes were made in the manuscript forms used to record the population. For example, the population schedule had an added column asking for the relationship of each person to the head of the household in which they lived, thus making clear important family relationships that could only be assumed in the three previous censuses. At the same time, however, the columns for the value of real and personal property were dropped from the form, thus eliminating the important measurement of total wealth. Numerous volumes of data compiled from the returns from all states were published in the early 1880s.
In 1887, very fortuitously for the census history of Texas, an agency of the state government conducted a new population count and in 1889 published the results as the First Annual Report of the Agricultural Bureau of the Department of Agriculture, Insurance, Statistics, and History, 1887‑8. Nearly a century later, cultural geographer Terry G. Jordan discovered this report and published an article in the Southwestern Historical Quarterly entitled “The Forgotten Texas State Census of 1887” (Volume LXXXV, April 1982). Jordan pointed out that the 1887 census contained county-by-county data on the ethnic makeup of the population of Texas that was unavailable in any other printed source and that similar data for the state in 1890 had been destroyed with the burning of the census of that year. In 2001, the Texas State Historical Association reprinted the entire report.
As previously discussed, the United States Census of 1890 was destroyed in a Commerce Department fire in the early 1920s, so no manuscript returns with data on individuals from that year are available. At least, however, more than ten volumes of compiled data, including statistics for Texas, were published.
The State of Texas has not conducted a major count of its population since 1900, but the decennial United States censuses have provided mountains of statistical data on Texans throughout the twentieth century and into the twenty-first. To protect the privacy of individuals, the manuscript population schedules are not made available to researchers until seventy years from the date the enumeration was completed. For example, the Census of 1950 is scheduled to be opened for researchers in 2020. However, the Census Bureau published compilations of the data as soon as possible after completion of the enumeration. From 1900 through 1940, these aggregated numbers were published in large volumes, each containing information on more than one state and its counties. Then, beginning in 1950, there is a volume for each state entitled Characteristics of Population. Texas was volume 43 in 1950 and became volume 45 in 1960. These volumes are divided into several sections such as "General Population Characteristics" and "General Social and Economic Characteristics."
The decennial censuses of the United States that began in Texas in 1850 have also included statistics of a non-population nature that, although technically not censuses, should be considered as important complements to the population returns. These include censuses of agriculture, manufactures, mineral resources, and business patterns.
United States censuses of agriculture were taken in Texas every ten years from 1850 through 1950. Special agricultural censuses were also conducted in 1925, 1935, and 1945. After 1950, agricultural censuses were taken in 1954, 1959, 1964, 1974, and 1978. Beginning in 1982, agricultural censuses were taken at five‑year intervals in years ending in 2 and 7. Data from the census of agriculture have been published by state and by county since 1850. Since 1945 there has been a volume on Texas alone. The census of agriculture contains statistics on the number and value of farms, livestock, and crops.
The census of manufactures was taken every ten years from 1850 through 1940. After 1940 this census was taken in 1947, 1954, 1958, 1963, and 1967. Beginning in 1972, it was conducted at five‑year intervals in years ending in 2 and 7. It contains data on employment and production in each of the industries found in each county.
Data on mineral industries were included in the census of manufactures or related volumes through 1940. Separate censuses of mineral industries were taken in 1947, 1954, 1958, 1963, and 1967. Beginning in 1972, this census was conducted every five years in years ending in 2 and 7. Obviously data from these censuses on oil and gas production are vital to the story of many Texas counties.
The publication of County Business Patterns began in 1946. This census was taken at three‑year intervals until 1964, and it has been annual since then. It provides county‑by‑county data on employment and payrolls in each type of business activity such as manufacturing, retail sales, and wholesale sales.
In summary, census data provide the basic framing for any local or state history. Texas researchers are blessed with numerous population counts that are accompanied by invaluable economic and social statistics. One of the most useful published histories of the census is Measuring America: The Decennial Censuses From 1790 to 2000, U.S. Census Bureau, 2002.