Of course, some of the suggestions for new questions were of the tongue-in-cheek variety. One writer for a South Carolina newspaper suggested that the 1930 census "neglected to put in many questions that the present age demands. . . . How much did you lose in the stock market crash last fall? If you could live your life over, would you get married? What is your average golf score? If single, how much do you spend on girls a year? Are you a native born or naturalized Democrat? How do you pronounce the name 'Aristide Briand?'"18

To make room for the new questions, the inquiries on the mortgage status of owned homes, date of naturalization, and native language of mother and father were dropped, and two questions on the ability to read and to write were combined into one.19

To alert the population to the coming census, the Census Bureau began designing its publicity program in 1928.20 This program began with efforts to inform the public about the general purpose for conducting the census as well as an explanation of the types of questions to be asked. Such information came to the public through a series of press releases and articles intended for publication in newspapers and magazines. In federal buildings, a presidential proclamation was distributed and displayed that stressed each individual's duty to respond to all questions; informed people that the census had no relationship to taxation, military conscription, jury duty, or compulsory school attendance; and reminded respondents that the information they provided would be held in strict confidence. And with the final publication of the printed reports, over a year after Census Day, the Census Bureau released a second set of press releases designed to highlight and explain the significance of the major findings of the census in both tables and text.

The Census Bureau, however, did not rely solely on the printed media to mobilize the public. Perhaps taking a cue from the "admen" of the private sector, the Census Bureau worked to appeal to the public in a variety of ways. In addition to asking local newspapers to announce the names of those selected to be supervisors and enumerators, it sought the cooperation of radio stations and movie theaters in promoting the census. The Census Bureau also appealed to local chambers of commerce and other community organizations to encourage the members of their communities to complete the census.

Another useful resource for education, promotion of the census, and recruitment was in the schools. Beginning in late 1929, the Census Bureau designed an outreach program aimed at the National Education Association to urge teachers to explain the census to their students. Students would then write an essay that they would take home to show their parents. In addition, the Census Bureau encouraged teachers, substitutes, and advanced students to become enumerators.

The Enumeration

The area covered by the 1930 census included the 48 states, the District of Columbia, and the territories of Alaska, Hawaii, Puerto Rico, the Virgin Islands of the United States, Panama Canal Zone, Guam, and Samoa.21 Census Day for all areas was the same except in Alaska, where the census was taken as of October 1, 1929, for "climatic" reasons. Territories accounted for approximately 1,700 enumeration districts.22

The transition from preliminary planning and preparation to early census operations had begun by mid-1929, when the Census Bureau started to identify candidates for the 575 census supervisor positions. The agency received a waiver from civil service laws for the hiring of supervisors and enumerators. This allowed the bureau to accept applications from political referrals as well as those from people without political credentials. Steuart reported that by June 1929, bureau headquarters was receiving applications for supervisory positions on a daily basis. Upon receiving an application, the agency sent the inquirer the fifty-page supervisor's manual with a letter asking him or her to read the manual and inform the agency if they were still interested. Not surprisingly, some applicants withdrew their applications upon finding out what the duties involved. Among other responsibilities, each supervisor would be responsible for overseeing the activities of an average of 170 enumerators.23

Applicants who survived this part of the process were sent an application blank, a test census schedule, and a descriptive narrative of a small population. They were to complete the application and the census form with facts contained in the narrative and mail the formal application package to bureau headquarters. There, agency officials corrected the test and sent a photostat of the corrected version to the applicant. If the person passed the test and could demonstrate having had business training or other qualifications for supervisory work, he or she was offered the position. Supervisors were expected to live in or near the city in which the supervisor's office was established. Thirty of the 575 supervisors appointed for 1930 were women.24

Once appointed, one of the supervisor's first responsibilities was to acquire office space and equipment, at no cost to the bureau wherever possible. Supervisors were instructed to seek assistance from the local Federal Business Administration and to request space, office equipment, and furniture from federal, state, and municipal agencies and from civic organizations such as chambers of commerce and boards of trade. Success in these endeavors varied widely. More than half the supervisors (346) were able to secure free office space, but less than a quarter (141) got free office equipment. In the balance of the offices, the agency was forced to rent space and rent or purchase the necessary equipment and furniture.25

Census supervisors were delegated the authority to hire and fire enumerators. Initially, all applications for enumerator positions were to be sent to bureau headquarters. Aspiring census takers were sent the same test census form, instructions, and narrative mailed to supervisors, as well as an application form, and asked to complete the census schedule using the information in the narrative according to the instructions. Enumerator packages were graded and corrected in Washington, and those that passed were sent to the appropriate supervisor. By March 1930, the number of enumerators needed and the tight time schedules led to an updated procedure in which applicants sent their completed packages directly to the supervisor's office, where they were graded and corrected.

Bureau officials estimated that they would need about 100,000 enumerators to take the census. In the end, the agency hired 87,800 enumerators. Of the nearly 198,000 people who applied for these positions, 157,000 were judged qualified.26

Enumeration districts in cities comprised an average of eighteen hundred people. Enumerators were expected to complete their work in the two weeks from April 2 to April 15. For rural areas, enumerators administered the agriculture census as well as the population census, and the districts contained fewer people. But because the area covered was much larger, rural enumerators were given four weeks to collect their data.

Enumerator pay varied in different parts of the country but was usually four or five cents for each person enumerated and forty or fifty cents for each farm. In exceptional circumstances, the rates went as high as twenty cents a person and five dollars a farm. The goal was to allow the average enumerator to earn between five and eight dollars a day during the enumeration.27

In addition to overseeing the work of the enumerators, census supervisors were also authorized to appoint an office staff. Usually hired in February or March 1930, the office staff consisted of two office supervisors, a stenographer, and up to ten clerical workers. Supervisory personnel earned between $4.00 and $6.00 a day; stenographers were paid $5.00 a day. Pay for clerical workers varied between $2.50 and $4.00 a day but could reach as much as $10.00 a day.28

Any number of factors could contribute to delays. In some parts of the Southwest and the far West, enumeration districts encompassed more than 150 square miles of rugged terrain. One enumerator working out of Butte, Montana, said that he had ridden six hundred miles on horseback between March 29 and April 23 in order to count the three hundred people in his district.29 In neighborhoods heavily populated by immigrants, suspicion of outsiders and the government in general made it virtually impossible for anyone but a local resident to obtain complete and accurate information. While the Census Bureau tried to hire locals in most cases, it was not always possible. While 291 translators were hired to assist with the enumeration, this was obviously an inadequate number for a foreign-born population of 14.2 million people.30 Despite these difficulties, most of the census data were collected within the time allotted.

After the office staff had reviewed the completed schedules returned by the enumerators, supervisors were told to release preliminary counts for as many civil divisions areas in their districts as possible. By mid-April, supervisors throughout the country had begun to release the figures. Two months later, preliminary counts for 745 cities and towns of 10,000 or more and for more than half the 3,098 counties had been released.31 Local officials were anxious to have these figures, and a few even complimented supervisors on the quality of the count in their areas. For example, the mayor of Santa Cruz, California, wrote to the local census supervisor to thank him for releasing the preliminary count for his city, adding "Your work should be commended."32

The official count of state populations for the purpose of apportionment were given to the secretary of commerce on November 17, 1930; he in turn sent them to President Hoover, who forwarded them to the Congress on December 4.33

Data Tabulation and Publication

Following a clerical review in the district supervisor's office, completed census schedules were sent to Census Bureau headquarters in Washington for further editing, coding, keying, and tabulation. Once the schedules were received at headquarters, they were reviewed for omissions and internal consistency. The next step in the process was adding the codes for occupation, birthplace, language, and other variables. Once the clerks supplied these codes, the schedules were ready for keying.

The tabulation system required that two cards be punched for each person. One card carried all the information noted on the individual census schedule, except age at marriage, which was included on the family card. The second, a family card, included the composition and characteristics of the family taken from the census schedule. A third card, the unemployment card, was prepared for each person who was usually employed but was not at work the day before the enumerator conducted the interview.34

The Census Bureau employed about two thousand people and up to 2,225 punching and verifying machines to keypunch over 139 million cards for the 1930 census of population.35

After keypunching, the cards were sent to be sorted and tabulated by machine. The sorting machines the bureau used could sort only one column at a time. While some tabulations were run without prior sorting, many of the card decks had to be sorted frequently to separate the subsets of information needed for a particular table. Groups of cards moved from the sorting machines to the tabulators.36 Unit counters (a kind of tabulating machine) added up the number of people in a given geographic area who had the specific characteristics called for in a particular table. Typical combinations included marital status by sex and race, and literacy by age. The end product of each tabulation run was a large sheet of paper with printed counts of the number of people in each of several groups defined by their characteristics. Each run covered a limited geographical area, such as an enumeration district. To obtain publication data, the results of two or more runs would be aggregated into publication geography, such as counties, cities, and minor civil divisions. As Secretary of Commerce Robert J. Lamont pointed out, each card had to be run through the electric tabulating machines "a dozen times or more" to extract the data presented in census publications.37 Nearly four thousand people worked in the sorting and tabulation area during the height of the 1930 census.