Every decade, from 1790 to the present, census marshals have canvassed the United States to record information on each person, farm, and manufacturing firm. The nature of modern census enumeration is infinitely more complex than that conducted in the nineteenth century but even by 1850, census recorders listed detailed information on the free and enslaved population, on agriculture, and manufacturing. Handwritten records of the responses to enumerators survive in manuscript form and on microfilm, where they are organized by state, county (parish in Louisiana), and then by civil districts (such as wards). The population schedule of the U.S. census lists free persons by household, the slave schedule enumerates slaveholders and the names and ages of their slaves, while the agricultural schedule details precise information on landownership, the cash value of farms, capital investment on estates, and the actual yield of each farm, by crop production, livestock etc. The manufacturing schedule lists industrial production. The 1850 and 1860 census included two additional schedules, persons who had died in the previous year and a final schedule which detailed information on taxes collected, schools, libraries, newspapers, church accommodations, pauperism, and crime for each locality. Collectively, the United States census is vast in its size and complexity. Documenting Louisiana Sugar provides users with the query functions to search the agricultural schedule for five key sugar producing parishes, 1850-1880. Farm-holdings are listed by individual name and thus can be readily traced, by those who wish to extend their search into the U.S. census, through internet census browsers such as ancestry.com

Despite its unquestionable value, the U.S. census is a problematic source to work with, no more so than in south Louisiana where the names and yields reported in Champomier and Bouchereau's sugar reports were often different to those listed by the federal census marshals. The Documenting Louisiana Sugar data entry team worked tirelessly to match as many records from the annual crop statements to the decennial census records. In most census years, we have been able to match approximately sixty percent of the data from the annual sugar statements to the appropriate census returns. That in its-self is a far higher match than might appear in the first instance; in fact, it is a singular achievement to have matched so many records and our thanks go to Alison Bambridge and the DLS project team for this laborious and incredibly time-consuming task. Indeed, as historians Mark Schmitz and Donald Schaefer observe, the diffuse and inconsistent nature of mid-nineteenth century census reporting frequently ensures that when sampling census data, researchers often return only a relatively narrow subset of entries, so problematic is the difficulty of matching census records. As Schmitz and Schaefer indicate, even matching census records from schedule one (population) and schedule four (agriculture) can be extraordinarily difficult; the DLS data team faced the added problem of matching the sporadic and often incorrectly filed census returns with the annual crop data. In so doing, we have attempted to match records by name (both grammatically and phonetically spelt), by location, by plantation, by yield (although these varied somewhat). Matching the DLS record set occupied the project team for approximately eight months so extensive was the task at hand, solely to match census and crop records for five sugar producing parishes for the years 1850-1880. Regrettably, we could not match the records for every sugar producing parish in Louisiana, but time and grant funds made this simply impracticable. Nevertheless, users of the DLS databases can consult the matched census and crop records for the key sugar-producing locations and also, of course, locate that record within the broader run of census materials.

Our difficulties were exacerbated by several fundamental problems associated with the U.S. census records. In the first instance, federal enumerators persistently undercounted the population and the number of cane producers in the state; moreover, they occasionally reported the name of the farm manager as the principal occupant of an estate that was owned by an absentee landlord. Estates owned and operated by partnerships in the Champomier and Bouchereau reports prove similarly problematic when linked to census records. Above all, the federal census marshals frequently misspelled names. The history of French, Spanish, and Anglo settlement in Louisiana, of course, ensured that the state maintained a large Francophone population throughout the nineteenth century. That population principally derived from French Acadian settlers who were expelled from Newfoundland in the 1750s and from French-Creole settlers who came from France and from the Francophone Caribbean during the eighteenth century, particularly following the Haitian revolution. German, Italian, and Chinese immigrants from the late 1860s onward further complicate the diversity of names and languages within the state. U.S. census enumerators frequently struggled with the linguistic variance present within their reporting districts and some made no attempt to spell names accurately, instead reporting phonetic spellings of French and Creole family names. To be sure, Champomier and Bouchereau were never as entirely accurate in their data entry and consistent in faithfully reporting family names as they themselves imagined or claimed, but their records far surpass in consistency (and reliability) the erratic nature of data collection and enumeration on behalf of the U.S. government's census bureau. Whereas Champomier and Bouchereau made professional careers from diligently recording names and yields, census enumerators sometimes lacked training and were poorly supervised. Moreover, they oftentimes lacked the diligence and conscientiousness that characterized the sugar reporters. The role of census enumerator in the mid-nineteenth century was considered political patronage and distributed with little thought for the particular suitability of the individual for the work at hand. Professional commitment to accurate census reporting thus varied (often considerably) as disparate enumerators took to the rivers and roads to chronicle the state's population and agricultural work. Although the forms supplied to them were standardized for data entry, the actual process of recording and enumerating census records was thus anything but regular, uniform, or consistent. From 1850, the Census Office in Washington oversaw data collection and attempted to introduce a degree of scientific discipline to the compilation and tabulation of data, but the manuscript census returns (which the DLS data entry team used on microfilm from the National Archives) remain very much a product of field research, with all the strengths and weaknesses of nineteenth century data collection.

Despite its shortcomings, the U.S. census is, of course, one of the only comprehensive record sets and time-series available for analysis. True, plantation records provide detailed and often complete production statistics (which we have drawn upon for our publications-see archival sources), but the census ultimately provides both micro and macro-level detail. It offers an invaluable portrait of agricultural production, every decade, and one that augments the infinitely more precise annual production data compiled by Champomier and the Bouchereaus. The manuscript agricultural census provides data from farm operators (free persons who operated farms which produced goods valued at more than $100 a year) and details farm size, farm value, and agricultural production. This information, considered accurate as of June 1 of the census year, was compiled in the manuscript schedule of the agricultural census. In contrast to the relatively Spartan nature of the population schedule, the agricultural schedule offered significantly more information than the other census schedules. For each farm operator, census-takers recorded data for 48 separate categories, or fields in 1850 and 1860, 53 in 1870 and 104 in 1880. In the sugar parishes, enumerators normally entered data in fifteen or so fields: name, farm size (acres of improved and unimproved land), the cash value of the farm and value of farm implements/machinery, the number and value of livestock, the agricultural produce of the farm (sugar, molasses, rice, Indian corn, peas, potatoes, dairy products, and cotton), and the value of the animals slaughtered. The 1870 census included provision for listing wages expended while the 1880 return offered columns to include the amount paid to laborers and the number of weeks hired workers were employed. Although the actual number of fields altered in the census records, enumerators in the sugar parishes focused most particularly on plantation crop output, farm values, capital investment, and farm acreage. Combined, these fields offer users a portrait of individual and parish-wide agricultural production, detail on specific farm operators, and a glimpse (albeit only every decade) into the land-use and labor employment patterns of the region's cane sugar producers. The value of farm implements allows one to calculate comparative rates of technical investment per improved acre between the sugar parishes and track the evolving pattern of production as the farms shifted from slave to free labor. By utilizing the cash value of farms, users can further examine the relative impact of civil war on the industry and examine the re-capitalization of Louisiana sugar by 1880. DLS users can, of course, track operators by name, but also order the parish census data by improved acreage, value of farm, sugar production, and cost of farm machinery.

Richard Follett

Sources: Mark D. Schmitz and Donald F. Schaefer, "Using Manuscript Census Samples to Interpret Southern Agriculture," Journal of Interdisciplinary History 17 (Autumn 1986), 399-414 (quotes 400-401); "Using the Manuscript Agricultural Census," Valley of the Shadow: Two Communities in the American Civil War, Virginia Center for Digital History, University of Virginia http://valley.vcdh.virginia.edu/Reference/censi/acensus.html