Skip to main content

The Great Enumeration

March 2024
19min read

“If we could first know where we are, and whither we are tending, “said Abraham Lincoln, “we could better judge what to do, and haw to do it. “For nearly two hundred years, the United States Census has been trying to find out.

Americans are a counting nation. They like figureslarge figures such as the gross national product, industrial production, consumer spending, consumption of energy, even measures of economic activity in such arcane areas as the production of brooms, brushes, and pickles. Especially do our people like to count themselves. This has been going on for a long time, serially in every year ending in zero since 1790. There is more to this than a mere quirk of national character. Statistics as an instrument for ensuring political equality are imbedded in the United States Constitution, which requires that political power be apportioned according to population. That is the primary historical and legal reason why we count ourselves every decade with a margin of error that has been thinned down to 2.5 per cent and is expected to fall still lower in 1980.

Since the beginning of recorded history, kings and potentates have numbered their subjects in order to tax them or to find out how many potential warriors their realms contained. There are indications of population counts in ancient Japan, China, Egypt, and in Babylonia; also in Greece, and among the Romans who gave us the word “census.” Hebrew enumerations are mentioned in the Bible. Monks were the principal custodians of vital statistics in the Middle Ages. In England’s American colonies, ministers, sextons, and elders kept church records which provided the heavenly demographic record of the elect and the damned. An increase in population was taken in New England as a sign of God’s favor toward the Congregational Church. Though no official government enumeration of the colonies as a whole was ever undertaken. But thirty-eight censuses of individual colonies were taken, most often when the Privy Council or the Board of Trade needed demographic facts for the governance of British America. The value of the information that was compiled varied greatly. The colonial governors, when commanded to initiate and supervise the task, could be energetic, or independent, lazy, overworked, or inclined to tell London what it wanted to hear. Local officials who did the actual work displayed little enthusiasm for so onerous an exertion which was, moreover, not part of their legally defined duties. The people, suspecting that the inquiries had something to do with taxes, were evasive and uncooperative.

It is the special distinction of the United States Census that it is regarded as being the first in modern times to conduct a periodic enumeration of the people and to be strictly uniform in its recurrence. But our census did not arise primarily from a sudden appreciation of its usefulness in the gathering of social statistics and economic information to guide public and private decision makers. It was, rather, the result of a masterly compromise, one of the great political achievements in our history. Because of it, the delegates to the Constitutional Convention of 1787 were able to complete their work successfully.

The background was this: under the weak Articles of Confederation the debts incurred by the central government for the common defense during the Revolutionary War were to be charged to the states in proportion to the value of their surveyed land, not their population. Congress consisted of a single house, every state having one vote. Congress could apportion the tax burden but each state retained the authority to levy and collect its share. This meant, in effect, that payment was voluntary; the United States could beg but not command it.

The adoption of the Constitution changed all that, but only after a monumental battle between the big states, which wanted representation in Congress to be determined by population, the small states, which wanted to retain the one-state, one-vote principle, and the Southern states, which wanted to include slaves whom the slaveowners regarded as people when they were thinking about their state’s representation in Congress but as property when they thought about taxation based upon population.

In the end the delegates worked out a most ingenious solution which made the Constitution finally acceptable to all: it called for a census to be taken within three years and every subsequent ten years, and Congress to consist of two branches—the Senate, where all states were represented equally, and the House of Representatives, where the populous states would have the greatest vote. In a concession to the Southern states, they were permitted to add to the number of free persons “three fifths of all other Persons,” meaning their slaves, for representation in the House.

The stroke of genius lay in a provision that direct taxes were to be apportioned in the same way as the House of Representatives. This tying together of representation and taxation removed any bias on the part of the states toward the census since it eliminated, as James Madison wrote in The Federalist , any temptation to cook the books, that is, “to swell or reduce the amount of their numbers. Were their share of representation alone to be governed by this rule, they would have an interest in exaggerating their inhabitants. Were the rule to decide their share of taxation alone, a contrary temptation would prevail. By extending the rule to both objects the states will have opposite interests, which will control and balance each other; and produce the requisite impartiality.”

 

In accordance with the constitutional requirement, the First Congress at its second session enacted a law governing the census. It was approved March 1, 1790, and within a year after George Washington became President the first enumeration was made. The work was done by the United States marshals of the several judicial districts with assistants of their choice, under the supervision of the Secretary of State, Thomas Jefferson. There were no standardized questionnaires, or “schedules,” as they are called in census parlance today, and the assistants, who were paid a pittance, had to provide their own pens and paper. (In fact, no printed schedules were furnished until the census of 1830.)

Though there were only six simple questions to be asked, the task was difficult because of poor transportation, uncertain boundaries, and a scattered population. The people were suspicious, like taxpayers from time immemorial, and consistently underreported. Some had never been counted before, and those good Christians who followed the Bible as a light to their feet feared any counting, and evaded the marshals and their aides when they could. For they remembered a most unpleasant account in the Old Testament of the sin of King David, who incurred the wrath of Heaven when he had his captains number the children of Israel, and so drew down a pestilence upon them. The 1790 census took eighteen months to complete and produced a head count of 3,929,214 Americans. The public, expecting a total above 4,000,000, was disappointed and there was anxiety about the political effect abroad. But President Washington explained the difficulties in detail and gave assurance that “our real numbers will exceed, greatly, the official returns.”

 

The original law mandating the census was continued with minor changes until 1850. Faced with the possibility of war with Great Britain, Congress in 1810 added an inventory of manufacturers to measure the industrial capacity of the country and its ability to fight or compete for new markets. More questions aiming to get at the resources of the nation, both human and material, were placed in the schedules in 1840. The enumeration was still carried out on a part-time basis by the U.S. marshals, and the information collected continued to be, as it always had been, open to the public. But the product was heavily criticized for its errors.

In answer to the complaints, under the 1850 census law Congress provided that raw data from the field must be sent to Washington to be centrally tabulated there. Further, instead of counting only the head of the family—the traditional method—each family member would be included separately, making possible more sophisticated analysis and cross-tabulation. In Studies in American Demography (1940), Walter F. Willcox called this the most important change in census history.

There was nothing especially notable about the eighth decennial census, but the ninth, taken in 1870, suffered from the disruptions of the Civil War, the disorganization of the government in the South, and the adjustment to the new status of blacks under the Fourteenth Amendment (this was the first census in American history that contained no questions about slaves). Spoilsmen controlled the work of taking the census in 1870. The U.S. marshals looked upon the census as an opportunity to distribute the offices to deserving henchmen, each of whom was instructed to “provide himself with a secure portable inkstand, good ink, and a sufficient number of pens. All entries will be carefully dried with the blotting paper which accompanies each portfolio.” The entire force of enumerators was taken from the Republican party, including some blacks who could not read or write. The able superintendent of the census, Francis Amasa Walker, had no control over the marshals or the quality of their work and was among the most severe critics of the census which he headed. There were newspaper attacks, too, arguing that the census should be confined to its narrow constitutional function; but Walker insisted: “What the country needs is more information, not less.”

Yet the pressure for more data consistently exceeded the capability of the technical and administrative machinery to provide and disgorge the answers. The encyclopedic censuses of 1880 and 1890 surpassed all previous efforts, employing more than two hundred schedules which asked an almost unbelievable thirteen thousand questions. They included, just for a few examples, the number of old soldiers and their widows still alive, the number of pounds of butter made, and data on fire losses, fire departments, garbage disposal, cemeteries, rail traffic, prisons, idiots, occupations, and unemployment. Besides this, the 1890 census attempted to include all Indians for the first time, which required a Solomonic decision as to who was an Indian. Biological, legal, and cultural considerations were taken into account. The Census Bureau decided that in addition to full-blooded Indians, persons of mixed blood were Indians if they were enrolled by a tribe or registered at an Indian agency, or if those who knew them regarded them as Indians. This definition undoubtedly admitted some Mexicans and whites who had dropped out of conventional American life. The information gathered included such fascinating facts as the number of Indian polygamists, the number of Indians killed during the year, including women and children, whether by soldiers or interested citizens, and the number of whisky sellers prosecuted.

Although the counts of 1880 and 1890 produced an acute case of statistical indigestion, they did profit from an exhaustive study of census operations conducted by a House Committee on the Ninth Census, headed by Ohio congressman James A. Garfield. The marshals were replaced by civilian supervisors, pay was liberalized, accuracy improved, and confidential information was protected by law. For the first time, women appeared, two hundred of them, among the army of fifty thousand census takers.

Up to and including the census of 1880, a multitude of clerks was employed in tabulating the census, shuffling the piles of schedules, making tally marks on ruled sheets, a task so slow and tedious that the summaries were out of date before they could be published. The outlook for the equally complex 1890 census was even more dismal when a fortunate meeting of two able minds led to the invention of tabulating machinery and the era of the punched card.

According to a generally accepted account, Dr. John Shaw Billings, a surgeon and managerial genius, then in charge of the Division of Vital Statistics of the 1880 Census, made a suggestion while walking through a room in which hundreds of clerks were bowed over their tables as they hand-tallied items, handling each schedule at least six times. Dr. Billings’ companion was Herman Hollerith, a young engineer who was preparing a report for the Census Bureau. Dr. Billings said to him, “There ought to be some mechanical way of doing this job, something on the principle of the Jacquard loom, perhaps, whereby holes on a card regulate the pattern.” Hollerith thought about the problem and ultimately produced a machine that could punch and sort cards on which information had been entered in code form. First used on a large scale in the census of 1890, the “Hollerith cards” immediately saved $5,000,000, two years in processing time, and set in train a sequence of developments which led to the modern computer.

Hollerith’s invention made possible the tabulation of complicated facts with such refinement that it became possible to identify (for example) a male, white, civil engineer, thirty-two years old, who had four years of college, who lived in Indianapolis (Tract No. 6, Enumeration District No. 1), was born in Pennsylvania, and had lived previously for five years in Ohio. The machine, moreover, could reject suspicious cards, such as the report of a person seven years old who was a physician and a widower.

As plans were made for the twelfth census, there was general appreciation of the fact that the last two had tried to do too much without time for careful preparation, since the organization was always dismantled after each census and had to be put together again every ten years. So this time the questions were greatly restricted in number, and many topics were covered by being transferred to a series of intercensal reports. The 1900 census is rated as one of the best, though one special agent who worked on manufacturing schedules in Philadelphia reported on many intractable problems in trying to decide who was a manufacturer, and in getting hard information from reluctant or naive small operators. He had to interrogate and interpret the answers of wage earners who moonlighted after supper repairing bicycles, housewives who took in sewing in September and October, and the owner of a factory who had to be cornered and asked, “How much would you have paid your sister, if she had been someone else’s sister?” About all that enumerators who recorded business figures could do around the turn of the century was reach an amicable agreement with the respondent on a set of figures that was highly speculative. But they did show reasonable internal relationships, item for item, and thus met “all the requirements of mathematical accuracy and statistical harmony,” if not of absolute truth.

In 1902 Congress at last passed legislation making the Census Bureau a permanent organization. Walter Willcox wrote an entertaining explanation of how this came about: “Director William R. Merriam handled Congress very cleverly; got a stunning group of girls on his staff; nearly all of them, no doubt, wanted to remain in Washington, and in the Census Office (at least until they got married). These girls, I was told, brought so much pressure on Congress that in 1902 … the office was made permanent, not for any scientific reason, but to keep the staff from being disbanded.”

For the first thirty years of this century, the Census Bureau proceeded with caution and limited its objectives, although there were further improvements in data processings A major innovation of the thirties, for instance, was the perfection of sampling, a technique based upon the theory of mathematical probability: finding a small group of individuals selected with precision, whose responses to questions provide an accurate profile for the population as a whole. By such means more information could be collected than before at reduced cost and without placing a great burden on respondents. Out of this work came a monthly report on unemployment, critically important during the New Deal, and still published as the Current Population Survey.

 
 
 
 
 
 
 

Self-enumeration was also tried, with poor results. Its time had not come. Immigration was still at full tide and a large population of new Americans had passed through our ports who could neither read nor write English. A demographic shift of historic proportions was recorded in 1920 and should not pass without notice. In that year the urban population became greater than the rural, amounting to about 51 per cent of the 106,000,000 Americans who were counted. Strangely enough, or rather by sheer accident, one of the most important events in U.S. history was not caught by the census: the Great Depression. This happened because the 1930 census was designed before the collapse, and by the next census the economy was rising to meet the demand for military hardware and our society was changed almost beyond recognition.

 

Some new questions m 1940 which the public was not prepared for raised a popular storm. Respondents were annoyed to be asked, “Do you have a toilet or privy?” or how much money they made. Information on salary and wages was badly needed, in the view of the government, to check on economic activity, current purchasing power, the extent of unemployment, the number of workers receiving substandard wages, and current hours of work. But Senator Charles W. Tobey denounced the questions relating to compensation both on the floor of the Senate and on the radio (more than 80 per cent of U.S. homes then had radios). The senator branded the inquiry as intolerable, un-American, unconstitutional, a violation of the Bill of Rights, and lacking legislative authority. Hearings were held on a resolution introduced by the New Hampshire senator, but the Senate leadership did not press for action on the floor.

Meanwhile, housewives in Olean, New York, adopted resolutions, organized a broom brigade, and threatened a march on Washington—where local merchants had launched an Anti-Snooping Club. Others who stood up to be counted against counting were the Defenders of America, the Fairfield (New Hampshire) Grange, Arthur Krock of the New York Times , the Sentinels of the Republic, Women Investors in America, the National Association of War Mothers, and a woman from Kenmore, New York, who said she knew of seven hundred women ready to go to jail rather than to reveal their incomes to the census taker. For a while it seemed that people might refuse to cooperate with the sixteenth census—though just those receiving five thousand dollars or more were asked only to say so; if they had other income all they had to say was whether it amounted to more than fifty dollars per year. Peace was restored when the objectors were permitted to report income information directly to Washington on a separate mail-in form. About two per cent of those answering chose this method.

The 1940 census also had to contend with a social phenomenon known as age heaping . This was the first population count taken after the passage of the Social Security Act. It turned up an extraordinary number of people who suddenly became sixty-five, the age of eligibility for Social Security benefits—many more than could reasonably have made it according to the demographic figures. It was also noticed that women in particular resisted entering the next decade of life. A large number were missing, for instance, from the agethirty group and peaked up, or heaped, at twenty-nine. Fifty-three-year-olds, it was also observed, strongly preferred a round number, like “fifty.”

By the 1940’s the punch-card system could no longer meet the demands of government agencies, large and small businesses, and the scholarly community: it was too slow. But a new electronic computer developed for the army suggested that a general-purpose digital computer might be designed to serve the needs of the Bureau of the Census by taking a problem, storing instructions, making lightningfast calculations, and printing out the answers.

After long and careful study by the National Bureau of Standards and a select scientific committee, the project was approved, a contract let, and a computer, with the trademark name UNIVAC I, was delivered in the spring of 1951. “These were wild times,” Ann Herbert Scott wrote in Census U.S.A. (1968), “when a small number of ‘computer-happy’ census staff worked night and day to master the new equipment.” For fourteen months UNIVACI was in continuous operation, twenty-four hours a day over a seven-day week, tabulating the data from the 1950 census. UNIVAC I was honorably retired in 1963 and some of its components may be seen in the Smithsonian Institution’s National Museum of History and Technology in Washington, but the bureau continues to rely upon new generations of ever-faster computers to do its awesome work.

While the early generations of computers attained dazzling internal speeds, they were handicapped by the need for faster devices to feed information into the machines and for faster ways to get the printed results out of the other end. These pressures in the late 1950's, with the 1960 census approaching, led to the development of a process with a folksy name, FOSDIC, which stands for Film Optical Sensing Device for Input to Computers. FOSDIC transferred data directly from microfilm to magnetic tape to be fed into the computer, and new high-speed printers spit it out for reading. In tabulating the 1960 census, the new system counted almost 180,000,000 people in a little more than three months, the findings occupying more than 200,000 printed pages. Manual copying, coding, and key punching were eliminated, along with their potential for such errors as were uncovered in the 1950 Census of Population. (Some years later two Princeton University professors had come upon a puzzling table in an obscure section of the tabulations. It showed that there were 565 fourteen-year-old widows, and 184 precocious Indian widowers and divorcés in the Northeast at the tender age of fourteen to twenty-four years. The explanation: a UNIVAC key-punch operator had hit the wrong key, a mistake known as “off-punching.”)

The use of computer technology in the late sixties sharpened public concern over possible invasion of privacy. In a court contest, United States v. Rickenbacker , the argument was put forward that detailed questioning about housing was in violation of the Fourth Amendment. The district court decided in favor of the government, the court of appeals sustained the decision, and the Supreme Court declined to review. Thus it was judicially determined that the gathering of household statistics was a legitimate function of modern government. (One housing question not asked in 1960 was whether the house had a kitchen sink. Living standards had risen to the point that the question was no longer considered necessary.)

During the 1960 count the Bureau of the Census made new experiments with the householders themselves completing the schedules. Results this time were so promising that the procedure was greatly extended in the next count. The chief reliance remained, however, on the door-to-door interviewers, the foot-slogging infantry of census work.

The Census Bureau started to plan for the ninteenth and most recent decennial census (1970) before the work on the eighteenth was completed. The challenge consisted of trying to find and count, as perfectly as human error will allow, the total of the inhabitants of the fifty states, the District of Columbia, Puerto Rico, Guam, the Virgin Islands, American Samoa, the Canal Zone, the Trust Territory of the Pacific Islands—every area of U.S. sovereignty or jurisdiction. Twenty-one pretests and dress rehearsals were held, measuring the effectiveness of questionnaire formats, the mailout/mail-back system of coverage, including the resolution of tough ethnic language problems, follow-up training for enumerators when there was no response to the mail-in form, the logistics of deploying temporary fieldworkers, the printing of millions of printed forms, and the organizing of a massive campaign of public information. At bureau headquarters in Suitland, Maryland, the run-through of FOSDIC-readable schedules on the latest model of the electronic magician demonstrated that FOSDIC could handle four hundred questionnaires a minute for feeding tape into the high-speed computer that then printed them out as statistical tables—good news for government at all levels, for private industry, for marketing experts and academics.

The bureau invited suggestions for questions from users of census material everywhere. Many proposals were rejected, among them questions on religion, union membership, the amount of taxes paid, smoking, child spacing, household pets, and Social Security. With all in readiness, on March 26,1970, President Richard Nixon issued a proclamation that the census would be taken beginning on April 1, and that “Every American can be sure that there will be no improper use of the information given. …”

Some people were not convinced. The inner-city residents of Trenton, New Jersey, and North Philadelphia, Pennsylvania, for example, were reluctant to open their doors or mouths to any caller representing the government, revealing their strong aversion to being listed in any government records. New Haven, Connecticut, respondents balked at answering a question as to “all persons who stayed here overnight on Tuesday.” A Wisconsin assemblyman urged his constituents to reject all questions other than name, address, sex, and marital status, and newspaper advertisements appearing in Madison just before Census Day urged noncompliance. There was also resistance in South Carolina. The prevailing mood there disturbed some enumerators enough to cause resignations. In fact, there was a national debate carried on in and out of Congress on individual privacy versus public need, and on the obligations which lay upon government after it came into possession of personal knowledge. In the end, however, the legal relationship of each citizen to the Census Bureau remained the same: answering the population and housing questions was mandatory.

Widespread publicity explaining the need for the data and emphasizing the confidentiality of the responses quieted down the incipient rebellion. Five court cases went to trial, but in the end a high level of compliance was obtained throughout the nation. Federal law states that information furnished to the Census Bureau can be used for statistical purposes only and not in any manner that could lead to the discovery of the identity of any person or business firm. Employees of the bureau are under oath to obey the Constitution and the law. They can be fined one thousand dollars for every violation and sent to prison for up to two years, but no cases have been proved or even prosecuted under the present statute governing the census.

There is one exception to absolute confidentiality. An individual who wishes to prove that he was born, and when and where, can for a small fee get the evidence from the bureau. Historians, genealogists, all who are interested in family history, have free access to the first ten censuses (1790-1880), except those few state tallies which no longer exist. The records for 1900 are available for restricted use. The 1890 schedules were destroyed by fire. Those for 1910 and after are closed, though Congress has not said that census confidentiality as stated in Title 13, United States Code, runs in perpetuity. Yet it is extremely unlikely that Congress would ever amend Title 13 except to make historical census data available to scholars at some remote future time. The Bureau of the Census thinks a century would be about right.

The final population count for 1970 was 203,184,772, and this great undertaking produced two thousand separate reports—as well as some responses that revealed changing social attitudes. One woman wrote to the Secretary of Commerce to complain that a married woman was required to identify herself as “wife of the head” though there was no box for “husband of the head.” In the next census the question as to who is “head” will be deleted. James Madison had proposed questions about occupations in 1790 but the suggestion was not acted upon at that time. But by 1970 the focus was upon the economic life of the nation, and twenty-three thousand occupations and nineteen thousand different industries had been classified by the Census Bureau.

Citizens of voting age have been moving around a lot since 1970, making the apportionment for representation in Congress, as well as that of local and state legislatures, out of date. Blacks are migrating away from the Northeast, many returning to the South. Texas, Oklahoma, and Louisiana are gaining rapidly in population. Florida and Arizona are gaining, too, but at a slower rate. Six congressional districts now have more voters than the average district has people. The census of 1980 will adjust these matters, and not least among its functions will be providing new guidelines for the distribution of billions of dollars annually to states, counties, and cities through federal revenue-sharing programs as well as helping local government entities to plan new schools and to locate transportation systems, housing programs, daycare centers, and job-training centers.

A considerable change in the political landscape of the House may be expected. Losers, if present estimates hold, will be Michigan, New York, Pennsylvania, Illinois, Ohio, South Dakota. Winners will be Arizona, California, Florida, Oregon, Tennessee, Texas, and Utah. By constitutional mandate, the census results will go to the President by January 1,1981, then to the House of Representatives, which will reapportion its 435 seats and direct the state legislatures to redraw congressional district lines according to population shifts and in keeping with the Supreme Court’s 1965 one-man, one-vote ruling.

Preparations for the twentieth census are very nearly complete as this is written. In April, 1978, the doit-yourself mail-back system was tested in Richmond, Virginia, and in two adjacent counties. The door-to-door technique for use in sparsely settled regions was rehearsed in two counties in the southwest corner of Colorado. And in September a census was conducted south of Houston Street in the Borough of Manhattan, New York City, where special inner-city problems arise because of the mixture of racial and ethnic groups. These programs, known as the “1978 Dress Rehearsal Census,” simulated conditions the bureau would face in the next great national head count.

The work horse will be the mail-back system, but the foot canvassers will still be on duty—two hundred and twenty thousand of them directed by twenty-two thousand crew leaders—ready to call upon delinquent or tardy households as well as rural areas and “Special Places,” such as convents, nursing homes, and jails. Strenuous efforts will be made to avoid undercounting blacks, to find citizens who are on the lam, people engaged in illegal activities, human waifs, owners of 3,000,000 second homes, and nonconformists who are not interested in helping keep track of the world they have rejected.

The bureau will find that there are about 2,000,000 adults who are not married but share the same living quarters. The census people allow the respondents considerable latitude in describing their living arrangements and have gone out of their way to point out that there are a number of reasons for such ménages beside the one that first comes to mind. Arthur Norton, a Census Bureau population expert, put it this way: “Unrelated men and women who are sharing living quarters with someone of the opposite sex are not necessarily shacking up,” and added that the government does its best to stay out of people’s personal business. So the 1980 checklist, reflecting the marked extension of such living patterns, will provide for a discreet entry as “partner” or “roommate.” No question will be asked next time around about the ownership of a television set for the same reason there is no inquiry about a kitchen sink—everybody is assumed to have one.

Already it is possible to see beyond 1980. Legislation now on the books requires a national inventory in 1985 and every mid-decade thereafter, an idea first proposed to Congress by President U. S. Grant in 1872. The quinary count is not intended to duplicate the historic decennial census of the “zero” years in apportioning the House of Representatives but simply to bring up to date information on changing patterns and totals to guide the distribution of funds and to respond to the needs of private data users who require updated figures more often than once in ten years.

 

Even now the bureau, looking ahead, has a pretty good idea of some of the shifts and changes that will affect political power and the quality of life in the early years of the country’s third century; it is just as well, for no matter how complex the world becomes, the mission of the census remains as simple and as profound as it has been for almost two hundred years: to find out how things are going in America.

We hope you enjoy our work.

Please support this magazine of trusted historical writing, now in its 75th year, and the volunteers that sustain it with a donation to American Heritage.

Donate