Gambar halaman
PDF
ePub

The information collected by the two schedules was collected, however, by the same enumerator, at the same time as those for the population schedule.

Use of Sampling in the 1940 Census. The 1940 sample was a representative cross-section of the entire population. Tabulations made from the sample would be as nearly as possible the same as if information concerning every person had been obtained. The sample enlarged the scope of the census and facilitated tabulations in the following ways:

Since the supplementary questions were asked only 1/20th (5-percent sample) as often as they would have in a complete census, the speed of field work was increased and carried out at a reduced cost, thus making it possible to carry more questions on the schedule.

• Tabulations based on the sample could be completed months ahead of the regular tabulations prepared from the general population-an especially important feature in times of national emergency and for obtaining quick preliminary counts of the distribution of the labor force by area, sex, age, etc.

The reduced cost of sample tabulations permitted the publication of data that otherwise would not be possible.

• Sample cards could be stored for subsequent tabulations not feasible for the entire population as the need arose. Sampling helped to adapt the census to newly developed needs, and to maintain continuity from one census to another.

Participants were selected for the sample by designating 2 of the 40 lines on each side of the schedule as sample lines and instructing the enumerators to ask the supplementary questions for each person whose name happened to fall on these lines. This method resulted in a 5-percent sample of all the lines in each geographic area. The actual percentage of persons drawn from the sample from any district would vary by chance, depending on how the names happened to "line up" as the enumerators proceeded with their enumeration.

The Seventeenth Census: 1950

As in 1930 and 1940, the 1950 Census was conducted according to the terms of the Fifteenth Census Act. The enumeration began on April 1, 1950, with 90 percent of the population having been enumerated by the end of the month (weather delayed enumeration in some areas until mid-May). All but 1 percent of the population had been enumerated by the end of June 1950.

The 1950 census encompassed the continental United
States, the Territories of Alaska and Hawaii, American
Samoa, the Canal Zone, Guam, Puerto Rico, the Virgin
Islands of the United States, and some of the smaller

136

islands and island groups within the United States' possession. 28 The census also made special provisions for the enumeration of American citizens living abroad (and their dependents), including the armed forces of the United States, employees of the United States Government, and the crews of vessels in the American Merchant Marine at sea or in foreign ports.

The census of Americans living abroad was attempted through cooperative arrangements with the Department of Defense, the Department of State, the United States Maritime Administration, and other federal agencies concerned. These agencies took the responsibility for the distribution and collection of specially designed census questionnaires for individuals and households. Other persons living abroad were to be reported by their families or neighbors in the United States; however, the quality of these data was considered suspect and they were not included in the published statistics.

Procedures to improve coverage. Several aids were employed to improve the completeness of the 1950 Census coverage. The most prominent were as follows:

A longer and better planned period of training was provided for enumerators.

■ Each enumerator was furnished with a map of his enumeration district, showing the boundaries of the area for which they were responsible.

An infant card had to be completed for each baby born after January 1, 1950 (since experience had shown that babies are easily missed).

A crew leader was assigned to supervise each group of approximately 15 enumerators. The crew leader was responsible for helping enumerators with "problem cases" and for spot-checking a sample of the dwelling units assigned to them.

A special enumeration of persons in hotels, tourist courts, and other places where transients usually pay for quarters was made the night of April 11, 1950. "Missed Person" forms were published in newspapers at the end of the field canvassing operations so persons who thought they had been missed could complete a form and mail it to the district supervisors.

• District supervisors made preliminary announcements of the population counted so that any complaints concerning the completeness of the enumeration could be submitted before the field offices were closed. If the evidence indicated an appreciable undercount, a re-enumeration of the area was conducted.

28 Although some smaller islands and island groups did not participate in the census, data for their populations were collected from other sources and included in the 1950 census.

Measuring America

■ Rates of population change were studied to evaluate the enumeration's completeness.

■ Vital and immigration statistics were used in conjunction with census data. (Since the population at a given census should represent the population at the previous census, with additions and subtractions resulting from births, deaths, and immigration, it is possible to calculate the expected population on a given census date and compare the actual total received.)

Following these procedures improved the coverage of the 1950 census over that of the 1940 census.29 (The components of population change were probably estimated more accurately during the 1940s than for the 1930s because not all states were consistently registering births and deaths until 1933.)

Post-Enumeration Survey. The 1950 census was further checked using a post-enumeration survey, in which a re-enumeration, on a sample basis, was conducted. The Census Bureau recanvassed a probability sample of about 3,500 small areas and compared these to the original census listings to identify households omitted from the enumeration. In addition to the check for omitted households, a sample of about 22,000 households was reinterviewed to determine the number of persons omitted in cases where the household had been included.

The Post-Enumeration Survey interviewers were given intensive training and supervision. Efforts were made to limit respondents to the person who was presumably best informed regarding the information desired, i.e., the person themselves. These precautions resulted in an expense per case in the Post-Enumeration Survey many times that of the original enumeration, and affordable only on a samplebasis.

The Eighteenth Census: 1960

The 1960 census began on April 1, 1960, in accordance with the requirements of an act of August 31, 1954 (amended August 1957), which codified Title 13 of the United States Code. By mid-April, 85 percent of the population of the United States had been enumerated with the count up to 98 percent by the end of the month. Several notable changes were made in the procedures for taking and tabulating the census. These changes were: 1) the greater use of sampling, 2) the development of procedures enabling most householders an opportunity to consult other members and available records when completing

29 For the decade 1930 to 1940, application of these methods suggests that the total net number of persons missed in the 1940 Census may have been about 1,300,000 more than that missed in 1930.

Measuring America

the questionnaire for their families, and 3) the use of electronic equipment for nearly all data processing work. 30 Sampling. In the 1960 census, a 25-percent sample was used. The greater use of sampling meant that the totals for some of the smaller areas were subject to a moderate amount of sampling variation, the usefulness of the statistics was not significantly impaired. Using a 25-percent sample of households eliminated nearly 75 percent of the processing expenses otherwise required for the items in the sample.

Enumeration procedures. The 1960 enumeration was divided into two stages-the first concentrating on quick coverage of the population and the collection of a few items for every person and dwelling unit, and the second devoted to the collection of the more detailed economic and social information required for sample households and dwelling units. Both stages used questionnaires left at the residence. to be filled out by one or more members of the family. The enumeration began prior to April 1, 1960, when an advance census form was delivered by the U.S. Postal Service to each household. The time between delivery of the form and the arrival of an enumerator to collect the household's information allowed the household to assemble information needed to respond to the census inquiries. Shortly after April 1, 1960, the second stage of the enumeration began. Enumerators made their rounds to collect the census data and left an additional form-containing the sample inquiries-at every fourth house visited. Households receiving the sample form were asked to complete the form and mail it to their local census office in the postage-paid envelope provided by the enumerator. When these mailed questionnaires were received at the census office, Census Bureau personnel checked the sample forms for accuracy and conducted telephone or personal inquiries to complete unanswered inquiries when necessary.

This two-stage enumeration was believed to be advantageous in that, in the past, enumerators were given only brief special training and were burdened with more instructions and work than they could effectively manage. By creating a two-stage enumeration the field work and training were reduced. Approximately one-third of the enumerator work force was retained for work in the second stage-receiving additional training that focused solely upon the content of the sample questions.

In specified areas (about 15 percent of the total population, characterized as living in areas of low population density and/or having inferior road networks), the two staged enumeration was combined, so that the enumerator collected and recorded sample data in the same interview in which the 100-percent inquiries were recorded.

30A. Ross Eckler, "Plans for the 18th Decennial Census," presented at the annual meeting of the Rural Sociological Society, Cornell University, August 27, 1959. Pp. 3-6. Morris H. Hansen, "Procedures for the 1960 Census of Population and Housing," presented at the annual meeting of the American Statistical Association, Chicago, II, December 1958.

137

The Nineteenth Census: 1970

When planning for the 1970 census, the need for an accurate count of the population was even greater than in the past because of the increasing tendency for governmental bodies to use population as a basis for distributing funds, and the more general awareness by local government officials and others of the potential effects of census undercounts.

Throughout the 1960s, researchers had reported that the population was increasingly resistant to the census. Studies had shown more alienation and distrust of government, and there appeared to be more organized attempts to protest the census. Furthermore, undercounts following the 1950 and 1960 censuses were blamed upon the enumerators' failure to follow instructions. Hence, stress was placed on simplified procedures, training, and quality control. Analysis of the results of the 1960 evaluation program and studies performed in the 1950s and 1960s indicated that the reasons for the undercounts were more complex. In particular, a substantial part of the undercount appeared to be due to either deliberate attempts by some segments of the population to be omitted from the census or the fact that they did not fit into any households by the "conventional rules" of residence. Even where the undercount was due to complete households being missed, the causes were frequently such that additional enumerator training produced only marginal gains.

This analysis led to a two-phase approach to coverage for the 1970 census. The first phase was the use of a basic census methodology that permitted knowledgeable outside sources to have an offer input into the list of housing units established by the census, and provided for automatic checks that enumerators actually completed a questionnaire for all known units.31 This was done in areas containing about 60 percent of the population through the creation of an address register independent of the enumeration phase, correction and updating of the register by U.S. Post Office employees familiar with their routes, and checks by Census Bureau employees to ensure that all housing units on the address register were accounted for when enumerators had completed their assignments.

A self-enumeration questionnaire was used in 1970 (as in 1960 for 60 percent of the population). Such questionnaires were believed to provide better reporting within households, because they provided respondents uniform census definitions and rules to follow for unusual household residence situations. In the areas containing the remaining 40

31 Throughout the census history, a small percentage of enumerators completed questionnaires by "curb stoning." Curb stoning meant the enumerator completed questionnaires for an individual or multiple households from the curb, without actually conducting an interview or checking the accuracy of their "guesses." This practice was motivated, in part, by the requirement to meet quotas or payment for work done on a "piece-of-work" basis.

138

percent of the population, more conventional listing procedures were followed, but with self-enumeration features.

The second phase of the 1970 enumeration was to superimpose on the regular census procedures projects specifically designed to increase coverage. Prior to 1970, studies of the effectiveness of a variety of devices for improving coverage were made, generally as part of large-scale tests conducted during the 1960s, which resulted in several coverage improvement initiatives.

The 1970 coverage improvement program included measures to improve coverage by (1) developing a more favorable public view of the census; (2) increasing the public's understanding of the importance of the census and its confidentiality; and (3) improving the enumerators' performance in hard-to-enumerate areas through intensive training and supervision. The specific changes made includedA sharp reduction in the number of questions to be asked of households-the number of inquiries on the questionnaire intended for 3 million households had been reduced from 66 to 23.

[ocr errors]

Questions on the adequacy of kitchen and bathroom facilities were reworded to remove any implication that the federal government was trying to ascertain with whom these were shared.

■ The Secretary of Commerce increased his supervision of the census and retained independent experts as census advisors.

■ A letter accompanied the census questionnaire that explained the need for data requested and emphasized the confidentiality of responses.32

Census questionnaires with instruction sheets were delivered by the U.S. Post Office to every household several days prior to "Census Day"-April 1, 1970. In areas with comparatively large populations of Spanish-speaking households, a Spanish-language version of the instruction sheet also was enclosed. Households either received a short-form questionnaire, which contained questions asked of 100 percent of the population (80 percent of the population received this form), or a long-form questionnaire, sent to 20 percent of the population, containing questions asked of 15 and 5 percent of the population.

In larger metropolitan areas and some adjacent counties (approximately 60 percent of the United States' population), households were asked to complete and return the questionnaire by mail on April 1, 1970 (resulting in an 87 percent mailback response rate), which was then reviewed by an enumerator or census clerk. Telephone or personal follow-up was made to complete or correct missing, incomplete, or inconsistent questionnaires. For the remaining 40

32 United States Department of Commerce News, April 18, 1969. P.1.

Measuring America

percent of the United States' population, instructions asked that the householder complete the form and hold it for pick up by an enumerator.

The Twentieth Census: 1980

For most of the United States, "Census Day" for the 1980 enumeration was April 1, 1980.33 As in past censuses, all questionnaires were to be completed giving information as of that date, regardless of when the form was actually completed.

The 1980 census also included two small surveys-the Components of Inventory Change Survey, which obtained information on counts and characteristics of the housing units that changed or stayed the same between 1973 and 1980; and the Residential Finance Survey, requesting data on mortgages, shelter costs, selected housing characteristics, and owner characteristics.

The use of a mailout/mailback questionnaire in 1970 had proven successful, and eased the follow-up operation burden. Furthermore, tests during the 1970 census indicated the feasibility of administering a mailout/mailback census in rural areas and small towns. As a result, the mail census areas for 1980 covered 95.5 percent of the United States population.

Field Enumeration. The 1980 field enumeration procedures were similar to those used in 1970, with the exception of the greatly expanded use of the mail for questionnaire delivery and return. Households received a questionnaire in the mail, completed it, and mailed it back to their local census district office. In those areas enumerated conventionally (i.e., through enumerator visits to the housing unit), the U.S. Postal Service delivered a questionnaire to each household 4 days prior to Census Day. Respondents were instructed to complete their questionnaires, but hold them until an enumerator visited the household. The enumerators collected the completed short-form questionnaires or helped the head of the household complete the form at the time of the visit, or completed a long-form questionnaire at designated housing units. Enumerators also enumerated individuals living in group quarters.

Publicity. The 1980 census incorporated an extensive advertising and promotion campaign. The focus of the campaign was to increase public awareness and cooperation with the census, i.e., to encourage households to fill out their census forms, and in mail census areas, mail them back to their census district offices.

33"Census Day" in northern and western Alaska was January 22, 1980, so the enumeration would be completed prior to the Spring thaw. As part of an agreement with the local governments, Census Day in the Trust Territory of the Pacific Islands (excluding the commonwealth of Northern Mariana Islands) was September 15, 1980, so teachers could be used as enumerators.

Measuring America

The campaign was directed by the Census Bureau's Census Promotion Office (CPO), established in the Summer 1978. The CPO secured the free services of the Advertising Council in directing the advertising campaign. The Council, in turn, hired the firm of Ogilvy & Mather to develop the campaign.

The promotion campaign incorporated media advertising, the distribution of information kits to magazines and newspapers and census promotional kits to over 100,000 schools, and the development of an extensive network of partnerships with corporations and private organizations interested in supporting the census. In addition, public relations specialists in the Census Bureau's regional and district offices handled a variety of more localized promotional activities, including obtaining time for public service announcements (PSAs) from local broadcast outlets, advising census managers on working with the press, partnering with local companies, and serving as liaisons with complete-count committees (over 4,000 complete count committees were organized throughout the country in an effort to generate local publicity and support for the census).

The Twenty-First Census: 1990

The twenty-first census of the United States was taken as of April 1, 1990. The census covered the 50 states, the District of Columbia, Puerto Rico, the Virgin Islands, and the Pacific Island territories (American Samoa, the Commonwealth of the Northern Mariana Islands, Guam, and by special agreement, the Republic of Palau.)

The 1990 census used two questionnaires-a short-form containing questions asked of the entire population and a long-form with additional population and housing questions asked of approximately 1-in-6 households.

The content of the 1990 census questionnaire was similar to that for 1980. The short-form questionnaire for households contained the items to be asked of all persons and for housing units. Those items-plus the population and housing questions to be asked on a sample basis-appeared on the long-form questionnaire.

For the 1990 census, the Census Bureau introduced the Topologically Integrated Geographic Encoding and Referencing (TIGER) system, which was developed by the Census Bureau and the U.S. Geological Survey (USGS). The TIGER system documented all streets, roads, rivers, lakes, railroads, and their attributes (names and address ranges, where appropriate), as well as the boundaries, names, and codes of all geographic entities used for data collection and tabulation for the entire United States, Puerto Rico, the Virgin Islands, and the Pacific Island territories. In addition to supporting the geocoding requirement, the TIGER system also provided a means to produce the many different maps required for data collection and tabulation.

139

With the mapping capabilities of the TIGER system and the use of a master address list, developed jointly between the Census Bureau and U.S. Postal Service, the Census Bureau mailed pre-addressed short- or long-form questionnaires to approximately 86.2 million households. The 1990 census questionnaire packages were mailed to households beginning in February 1990. Most post offices had delivered the initial mailings by March 23, 1990, followed by a mailed reminder card on March 30, 1990. Occupants were asked to complete these questionnaires and return them by mail. Nonrespondents to the questionnaire mailout received a personal visit from an enumerator seeking to complete a census questionnaire for the household.

Publicity. As in 1980, the 1990 census was extensively advertised in television, radio, print, and public advertising. The goal of this advertising was to encourage mail response, reduce differential undercount, and foster a positive atmosphere within which to take the census convincing people that the census was both important and safe.

Promotion activities included "complete count" committees, information kits for schools, churches (the Religious Organizations Project), and the media, workshops, "pro bono" PSAs sponsored by the Advertising Council; local government outreach and partnerships; and the Census Education Project (designed to educate students about the census).

Compared to the estimated $38 million worth of free commercial advertising received in 1980, an audit placed the 1990 figure at about $66.5 million. Local television and radio stations were responsible for 69 percent of that value, followed by 21 percent for ethnic media. The PSA campaign reached a potential audience of 99 percent of the adultaged population, with an average of 68 exposures to census related advertising per person.

The Twenty-Second Census: 2000

The twenty-second decennial census-Census 2000 enumerated the residents of the United States, Puerto Rico, and the Island Areas, and Federal employees and their dependents living overseas as of Census Day, April 1, 2000. The majority of these households participated in the census through a mailout/mailback operation.34

34 Puerto Rico was enumerated using Update/Leave methodology-enumerators personally delivered a questionnaire to each household, after which the household completed the questionnaire and mailed it back to the Census Bureau.

In February 2000, the Census Bureau mailed advance letters to each household within the U.S. informing them that a Census 2000 questionnaire would soon be arriving. The letter also included instructions on how to obtain an in-language questionnaire for non-English speaking households.

Beginning in early March 2000, the U.S. Postal Service began delivering approximately 98 million questionnaires to households throughout the U.S. and its territories. The majority of households (83 percent) received a short-form questionnaire that asked for information on seven subjects (name, sex, age, relationship, Hispanic origin, and race).35 A sample of 1-in-6 households (17 percent) was selected to receive the long-form questionnaire, which in addition to the short-form questionnaire inquires, also contained 52 questions requesting more detailed information about housing, social, and economic characteristics of the household. The questionnaire mailout was followed by the mailing of "reminder cards" to each household receiving a questionnaire.

In total, 65 percent of households responded to the mailout/mailback census. The remaining 35 percent of households were visited by enumerators who attempted to complete a questionnaire via personal interviews.

Census 2000 Advertising Campaign. Census 2000 featured the first ever paid advertising campaign. So as to reach all adults living in the United States (including Puerto Rico and the Island areas), the Census Bureau awarded a contract to Young & Rubicam, totaling $167 million, for print, television, and radio advertising for its national, regional, and local advertising campaign.

The advertising campaign consisted of more than 250 TV, radio, print, outdoor, and Internet advertisements-in 17 languages-reaching 99 percent of all U.S. residents. By the end of the campaign, the census message "This is your future. Don't leave it blank."-had been heard or seen an average of 50 times per person. At its conclusion the campaign was ranked the second most effective campaign according to an AdTrack-USA TODAY consumer poll and for the first-half of 2000, the Census Bureau ranked 53rd in spending among all advertisers in the United States.

(For additional information on Census 2000 operations, see "Census 2000 Operational Plan," U.S. Department of Commerce, Economics and Statistics Administration, U.S. Census Bureau, December 2000.)

35For the first time, recipients of the short-form questionnaire had the option of providing their information by submitting electronic responses to the questionnaire via the Internet. Although this option was not extensively promoted, approximately 66,000 households chose to respond electronically.

140

Measuring America

« SebelumnyaLanjutkan »