AUSTIN, Texas -- On the first Monday of August in 1790 when an infant and independent America needed to gauge its military potential, the first U.S. census began. On horseback and on foot, U.S. marshals and their assistants rode and walked the dirt roads of a rural nation to record only the names of the heads of household and a head count of anyone else living there.
It began America's obsession with counting one of the fastest-growing and most heterogeneous populations in the world. Since then, writes Margo Anderson, a professor of history at the University of Wisconsin at Milwaukee, "political commentators and historians have analyzed and debated, and praised or condemned, and wondered over the document that the framers produced."It was 1787 when delegates from the original 13 states met in Philadelphia to discuss amendments to the Articles of Confederation. From the discussion came the order to take a census every 10 years of all "persons ... excluding Indians not taxed," for the purpose of apportioning seats in the House of Representatives and assessing federal taxes. (Indians not taxed were those who didn't live in settled areas. Later, American Indians everywhere were considered part of the population, but they were not included in apportionment figures until 1940).
The Founding Fathers were at odds, however, about how to conduct a census. Should apportionment be based on land assessments? Other measures of wealth (money, livestock)? Or population? It quickly became evident that population was easier to measure than wealth.
The 650 marshals and their assistants, who had to supply their own paper and pens to record the first count, went house to house unannounced in 1790 and wrote down only the name of the head of the household, almost always a man. They did only a head count of all others living in the house.
Enumerators distinguished their counts between free white males 16 and older and free white males younger than 16. What the government really wanted to know was how many men were eligible for military service. They also counted all white females, all other free people regardless of race and gender, and slaves. Slaves, who were primarily blacks, were counted as three-fifths of a person for apportionment purposes.
The early censuses lacked continuity: After each count, the census office under the secretary of state disbanded. What Anderson calls "population enthusiasm" began after 1870, when people recognized the need for an office that did not have to start from scratch every decade. Hence, Congress created a permanent agency in 1902 under the Department of the Interior. A year later it was moved to the Department of Commerce and Labor.
The zest to be counted emerged right away, says Anderson, who has written "The American Census: A Social History" (Yale University Press, 1988), co-written "Who Counts? The Politics of Census-Taking in Contemporary America" (Russell Sage Foundation, 1999), and is considered an expert in the U.S. census process.
"The American logic is that I want to get counted because I mean something. At first, there was competition between regions for social and political dominance. Population enthusiasm is embedded in the political culture of our society. That's the rule of the game. More people equals more power," she says.
EARLY CENSUS-TAKING
Marshals' assistants were the backbone of early counts. They earned $1 for every 150 people counted in rural areas or $1 for every 300 people counted in towns and cities.
"In 1790, they didn't get paid all at once. They got paid a little in the beginning and the rest when all the census information was recorded," Anderson says.
U.S. marshals earned from $100 to $500 for overseeing the counting in their jurisdiction. Today's enumerators, who are temporary workers, fare better: The pay range is $8.25 an hour to $18.50 where competing wages are high.
That first census took 18 months to complete, compared with the nine-month deadline the U.S. Department of Commerce has to deliver its report to the president.
The work of the marshals' assistants was tough in the early days. The biggest obstacle was the frontier nature of America. The country was overwhelmingly rural. Workers had no maps to guide them. "Transportation and communication were slow," Anderson says. "It was a slow-moving enterprise."
And to ensure that people cooperated, anyone who refused to answer could be fined $20 -- $10 going to the marshals' assistants and the other half to the government. Still, people were suspicious of any government worker and particularly wary of anything that might mean being taxed.
In 1790, U.S. marshals were required to post complete census reports "in two of the most public places within (each jurisdiction), there to remain for the inspection of all concerned." Returns were collated in Secretary of State Thomas Jefferson's quarters in New York City. The cost was $45,000, and the count was 3.9 million people, with 202,000 living in cities and the rest on land where agriculture was the primary means of support. President George Washington and Jefferson were distressed that the count didn't exceed 4 million -- many wanted to impress Europeans with the new nation's muscle, and muscle meant people.
ADVENTURES IN COUNTING
We don't know much about the life of census takers in the 18th and 19th centuries, but during the 1960 census, enumerators took better notes. One woman told a census taker that she was 145 years old. A 38-year-old woman claimed to have a 33-year-old son. A census taker was greeted at an apartment by a man taking off his shirt. The surprised enumerator began backing away and asked the man what he was doing. He said: "Oh, I was showing you my incision. Aren't you the visiting nurse?"
Animals have provided great challenges to census takers, archive documents show. In California, a goat chased a worker away and ate some of the completed forms he was carrying. Another worker was chased by two angry bulls. When she climbed over the fence to safety, a goose on the other side bit her.
No enumerators have been killed in the line of duty. However, a census taker was shot in New Mexico, according to the archive, which doesn't give a date for the incident.
Despite some of the perils of going door to door, workers got a few laughs from people's answers. They listed occupations that included snake doctor, mouse farmer, plumber's helper's helper and professional mourner. One person who lined the inside of caskets said she was an interior decorator.
Some census takers got a unique perspective on American life: One female worker helped deliver a baby at a farm. When the worker asked the mother for the baby's name to record in the census form, the woman answered, "Quitten."
"Don't you mean Quentin," the census taker asked.
"No," the mother said, "This is our eighth child and we're naming him Quitten."
WHO IS ASKING?
A big misconception about the census is who asks the questions. It isn't the Census Bureau -- its charge is to conduct the count. Behind the questions are government agencies such as the departments of Transportation and Housing and Urban Development, which need the information for planning and running programs. Congress approves every question, and billions of dollars are at stake.
Unlike today, questions posed in early census-taking did not cause a stir. Through the early 1800s, the basic information was always asked: name, age, sex, race and number of slaves. The scope broadened in 1810, when inquiries about manufacturing were made. In 1830, the government wanted to know the number of deaf and blind people in households.
The "insane and idiotic" were added to the rolls in 1840. That same census asked for the number of working people in the household, their occupation and literacy. In 1850, questions about occupation and birthplace appeared. A decade later, the government wanted to gauge the value of real estate and school attendance. In 1870, Uncle Sam was curious about people who had been denied the right to vote and women were counted in the labor force tally. Freed blacks were identified by name and the first data on Chinese was collected. Addresses were recorded in 1880.
The turn of the century saw a huge interest in counting immigrants and whether they had been naturalized. In 1900, Indians were asked for tribal affiliation and whether they were full or mixed blood. By 1950, it was necessary to gather data about the number of hours worked in a week. By 1970, people of Hispanic or Spanish descent were asked to identify themselves in that manner.
Several milestones mark the history of the census. In 1850, an individual became the unit of enumeration rather than the family. Each person's name and information about them was recorded. Thus, the process became more detailed.
The first mechanical tallying machine was invented to count the 1880 census. Temporary workers, some of them women, replaced marshals and their assistants. In 1890, workers who tabulated results welcomed the introduction of punch cards that were inserted into primitive electric machines for tabulating purposes.
THE NEW MATH
The development of scientific sampling techniques in 1940 was monumental, says Dave Pemberton, a historian in the U.S. Census Office. Sampling is a way of selecting a subgroup in the population and using their answers to gauge the population.
"In the 1930s, there was only anecdotal evidence of how bad the Depression was," Pemberton says. "If the government was going to make efforts to find people jobs and housing, they needed a way to measure it. Scientific sampling, a branch of applied mathematics, was done more frequently. They didn't have to wait 10 years for the next census to gauge how programs were working."
Sampling remains controversial. Although some propose that it be used to get a better measure of traditionally undercounted groups, the Supreme Court ruled last year that it cannot be used in apportioning House seats.
One of the most significant advances came in 1951 when the Census Bureau bought the first civilian computer from University of Pennsylvania Professors J. Prespet Eckert and John Mauchly. The room-size UNIVAC was used to tabulate data.
"This was the beginning of automatic computing in the United States," Pemberton says.
The next biggest change came in 1960, when questionnaires were mailed to homes with the understanding that people would mail them back. Instead of an enumerator tallying people, the new way relied on one person in the household to provide information about everyone living there.
In 1990, the Census Bureau teamed with the U.S. Geological Survey to develop a large map database called the Geographic Information System, which allows the bureau to break the nation down to city blocks and school districts. The computer technology is beneficial for enumerators and local governments, which use the maps for planning purposes.
The 2000 census marked the first time for professional advertising, $167 million, to promote the head count. This was also the first time for the advance mail notices to homes that the questionnaire was coming.
Plans have already started for the 2010 census. And planners are talking about a more frequent American Community Survey to collect detailed information about social and economic factors that might eliminate the long form of the present decennial census.