American Community Survey (ACS)

In a decennial census, each person is to be enumerated as an inhabitant of his or her "usual residence." Usual residence is the place where the person lives and sleeps most of the time, or the place he or she considers to be his or her usual residence. This place is not necessarily the same as the person's legal or voting residence, nor did the census always count persons as residents of the place where they happened to be staying on Census Day.

The American Community Survey (ACS) concept of residence differs from this decennial census concept. Because the survey is conducted every month on independent samples, and produces annual or annual average estimates, a concept of current residence was adopted as being more appropriate than the usual residence concept of the decennial census. The census requires that everyone have only one usual residence, since its primary purpose is to provide counts for the apportionment of congress. The concept of current residence being used in the American Community Survey also requires that each person have only one residence at any point in time, but that residence does not have to be in the same place throughout the year. This concept allows the survey information to more closely reflect the actual characteristics of each area.

To implement the current residence concept, the "Two-Month" rule was established. This rule states that if a person is staying in a sample unit at the time of survey contact, and is staying there for more than two months, he or she is a current resident of that unit whether or not the unit is also the person’s usual residence under census rules. If a person who usually lives in the unit is away for more than two months at the time of survey contact, he or she is not a current resident of that unit. Anyone staying in the unit at the time of survey contact who has no other place where they usually stay is considered a resident of the unit. The time of survey contact is defined to be when the respondent completes the survey questionnaire, or when the unit is reached by telephone or through a personal visit during the followup for mail nonresponse.

In the vast majority of areas of the country, the use of usual residence or current residence as the classification basis would produce substantially the same statistics. There might be appreciable differences for areas where large numbers of people spend several months of the year in units that are not their primary residences; for example, Florida, Arizona, and in beach or mountain vacation areas with large seasonal populations.

Interviewing Rules

The Census Bureau classifies all living quarters as either housing units or group quarters. Housing units, whether occupied or vacant, were eligible for inclusion in the ACS since 1996. The Group Quarters population was sampled in the 1999 and 2001 ACS. Each person whose current residence was an ACS sample unit in one of our survey sites was to be included in the survey, without regard to the person's legal status or citizenship. Persons residing in housing units were excluded from the survey only if the residence rules defined their current residence to be somewhere other than the sample housing unit.

Residence Rules

For housing units, each person included in the survey must be a current resident of the unit in sample--a place where he or she is staying for more than two months at the time of survey contact. If a person has no place where he or she usually stays the person is to be considered a current resident of the sample unit regardless of the length of the current stay.

Persons away from their residence for two months or less, whether in the United States or overseas, on a vacation or on a business trip, are considered to still be "in residence" at the sample unit and the unit is classified as occupied. Persons away from their residence for more than two months are considered not to be in residence. If no one is in residence for more than two months, the unit is classified as vacant. Units occupied only by persons who are staying for two months or less and who have another more permanent residence are considered to be temporarily occupied units. As with vacant units, only the housing information is collected for temporarily occupied units, and their survey information is tabulated with that of seasonal vacant units.

The "Two-Month" rule, with few exceptions, determines the current residency of all persons. These exceptions are noted below.

Children Away at School--Children in boarding schools below the college level are considered residents of their parental home. College students’ current residency is established by the two-month rule.

Children in Joint Custody--Children who live under joint custody agreements and move often between the separate residences of their parents are considered to be current residents of the sample unit if they are staying there when contact with the unit is made.

"Commuter Workers"--Persons who stay in a residence close to their work and return regularly to another residence, usually with family, are considered to be current residents of the family residence, not the work-related one.

For Group Quarters (GQs), a person is considered a resident if he or she is a resident of the GQ at the time of the interview.

Data Collection Procedures

The ACS is conducted primarily through self-response. The questionnaire mailing packages include general information about the ACS, and an instruction guide explaining how to complete the questionnaire. Questionnaires and instruction guides are available in both English and Spanish.

Interviewing for Housing Units

Each housing unit in the independent monthly samples selected from all the survey sites is mailed a prenotice letter announcing the selection of the address to participate, a survey questionnaire package, and a reminder postcard.  These sample units receive a second (replacement) questionnaire package if the initial questionnaire has not been returned by a scheduled date. (In the 1996-1998 ACS, field representatives delivered questionnaires to sample units in Fulton County, PA because we did not have a mailing list.  These housing units received only a survey questionnaire package.)

The mail-out/mail-back procedure is used where mailing addresses consist of a house number and street name. The Census Bureau developed a Master Address File (MAF) for these areas that served as the housing unit sampling frame for the survey. For the 1996-1998 ACS, the MAF was constructed by matching the areas’ 1990 Address Control File addresses and United States Postal Service Delivery Sequence File.  Since 1999, the the ACS has sampled addresses from the MAF developed for Census 2000.

Follow-up

Nonresponse Follow-up

In the mail-out/mail-back sites, sample units for which a questionnaire is not returned in the mail and for which a telephone number is available are defined as the telephone nonresponse followup universe. Interviewers in the Census Bureau's telephone centers, using Computer Assisted Telephone Interviewing (CATI), attempt to contact and interview these mail nonresponse cases.

Sample units from all sites that are still unresponsive two months after the mailing of the survey questionnaires and directly after the completion of the CATI operation are subsampled at a rate of 1 in 3. The selected nonresponse units are assigned to Field Representatives, who visit the units, verify their existence or declare them nonexistent, determine their occupancy status, and conduct interviews using Computer Assisted Personal Interviewing (CAPI).

Coverage and Edit-Failure Follow-up

The ACS questionnaire, like the 1990 census long form, consists of many questions concerning the characteristics of a household’s residence and its members.

In 1996-1998, questionnaires returned by mail underwent an intensive clerical review. A telephone follow-up (TFU) operation was conducted on all those cases that did not meet the high quality standards because of incomplete information, showed evidence of possible confusion on the respondent’s part in deciding who was to be listed as a household member on the questionnaire, or indicated that the household consisted of more than five people. These households were contacted by telephone to clarify who was to be included in the household and to obtain the missing information. The information was entered onto the mail return questionnaire by the telephone clerk. For households with more than five persons, a "continuation form" consisting of only the population questions was completed and attached to its mail return form.

Since 1999, this TFU activity has been conducted using a separate TFU CATI system.  The answers that a respondent provides on the mailed form are keyed, run against a program that is the same as the clerical edit noted above, and then show up as answers in the TFU CATI version of the case.   An interviewer uses the TFU CATI system to ask questions and enter responses to questions that were previously blank.  Also, for households with more than five persons, the interviewer obtains all appropriate information for the persons in the household whose data was not on the original questionnaire.

Processing Procedures

Questionnaires are returned by mail to the Census Bureau’s processing center in Jeffersonville, Indiana.

In 1996-1998 the questionnaires were "checked-in", clerically edited, and followed up if necessary. Questionnaires that did not require follow-up were sent directly to the data entry unit to be keyed. Questionnaires that failed the clerical edit were forwarded to data entry after the completion of the telephone failed edit followup. Mail returns that arrived more than three months later than their designated sample panel month were not allowed to be checked in or processed.

Since 1999, questionnaires are "checked-in", keyed, and then sent for TFU, if necessary.

Files of data responses from the questionnaires are combined with the response records from the CATI and CAPI operations and are processed further through automated coding programs that assign codes for place-of-birth, migration, place-of-work, ancestry, language, race, and Hispanic origin. A clerical coding operation is used to assign the codes for industry and occupation. The coded information is merged with the precoded information from the data files, and a final data file produced for editing, weighting, and tabulating operations.

Adjusting Dollar Amounts

The data processing steps include the adjustment of all dollar amounts, based on the difference between the average Consumer Price Index (CPI) for the household's 12-month reference period and calendar year of the interview. For example, a household that is interviewed in July of 2003 would have an income reference period of July 2002-June 2003. Income estimates for this household are expressed in calendar-year 2003 dollars, based on the difference between the average CPI for July 2002-June 2003 and January 2003-December 2003.