–B–
Residence Concepts and Questions in Selected Foreign Censuses

In this appendix, we summarize general approaches to residence concepts used in foreign censuses, as well as specific questions and instructions used on their census forms. We begin with a description of two sets of guidelines promulgated by cross-national agencies before describing practices in individual nations. This analysis only covers select countries that continue to conduct traditional head-count censuses (as opposed to records-based censuses or rolling censuses that infer population counts from a sample), and is also limited to nations for which English-language versions or translations of the census instrument are available from the Internet.

B.1
UNITED NATIONS/ECONOMIC COMMISSION OF EUROPE GUIDELINES

In 1998, the United Nations and the Statistical Office of the European Communities jointly issued a set of suggested guidelines for population censuses (U.N. Economic Commission for Europe and Statistical Office of the European Communities, 1998; hereafter, UNECE, 1998). Those guidelines explicitly linked the definition of “usual residence” to sleeping: “A person’s usual residence should be that at which he/she spends most of his/her daily night-rest” (UNECE, 1998:10–11). However, the guidelines immediately list several caveats to that definition:



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census –B– Residence Concepts and Questions in Selected Foreign Censuses In this appendix, we summarize general approaches to residence concepts used in foreign censuses, as well as specific questions and instructions used on their census forms. We begin with a description of two sets of guidelines promulgated by cross-national agencies before describing practices in individual nations. This analysis only covers select countries that continue to conduct traditional head-count censuses (as opposed to records-based censuses or rolling censuses that infer population counts from a sample), and is also limited to nations for which English-language versions or translations of the census instrument are available from the Internet. B.1 UNITED NATIONS/ECONOMIC COMMISSION OF EUROPE GUIDELINES In 1998, the United Nations and the Statistical Office of the European Communities jointly issued a set of suggested guidelines for population censuses (U.N. Economic Commission for Europe and Statistical Office of the European Communities, 1998; hereafter, UNECE, 1998). Those guidelines explicitly linked the definition of “usual residence” to sleeping: “A person’s usual residence should be that at which he/she spends most of his/her daily night-rest” (UNECE, 1998:10–11). However, the guidelines immediately list several caveats to that definition:

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census …problems [in stating a usual residence] may arise in dealing with the following groups of persons: persons who maintain more than one residence, e.g., a town house and a country home; students who live in a school or university residence, as boarders in a household or as a one-person household for part of the year and elsewhere during vacations; persons who live away from their homes during the working week and return at weekends; persons in compulsory military service; members of the regular armed forces who live in a military barrack or camp but maintain a private residence elsewhere; persons who have been an inmate of a hospital, welfare institution, prison, etc., for a sufficiently long time to weaken their ties with their previous residence to which they may return eventually; persons who have been at the place where they are enumerated for some time but do not consider themselves to be residents of this place because they intend to return to their previous place of residence at some future time; persons who have recently moved into an area and may not feel that they have lived there long enough to claim it as their place of usual residence—this may apply in particular to immigrants from abroad; persons who have left the country temporarily but are expected to return after some time; …and nomads, homeless and roofless persons, vagrants and persons with no concept of a usual address. “The treatment of all these cases should be set out clearly in the census instructions,” note the guidelines, and, “if possible, objective rules should be formulated for dealing with them” (UNECE, 1998:11). The guidelines further suggest that “people in groups (a) to (i) should treat the address at which they spend the majority of their daily night-rest to be their usual residence. For persons with a spouse/partner and/or children, the usual residence should be that at which they spend the majority of the time with their family…. People in group (j) should be treated as usually resident where they are enumerated.” Of particular note is point (f) on the treatment of persons in “institutional households,” which the guidelines later define as “persons whose need for shelter or subsistence are being provided by an institution” (UNECE, 1998:42). The language of (f) evokes the “enduring ties” concept of Franklin v. Massachusetts (Box 2-5), but suggests that the tie can decay after a “sufficiently long time.” In the discussion of “usual residence,” the guidelines provide no specific guidance on what constitutes such a time period, but in

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census discussing “institutional household” the guidelines note that “people who are normally members of private households but who are living in institutions …are only considered members of institutional households if their absence from the private households exceeds the one-year time limit specified for the place of usual residence topic” (UNECE, 1998:42). B.2 UNITED NATIONS STATISTICS DIVISION The joint recommendations of UNECE (1998) with regard to usual residence are consistent with—albeit more detailed than—general guidance provided by the United Nations Statistics Division (1997) to member nations. Indeed, the Statistics Division guidance merely provides the literal definition of “place of usual residence” as “the geographical place where the enumerated person usually resides.” Acknowledging that some groups may have difficulty specifying a usual residence, the 1997 guidelines state only that “the treatment of all such cases should be clearly set forth in the census instructions.” A planning paper suggesting revisions to the 1997 principles acknowledges their deficiencies, noting in particular that “defining the place of usual residence as the geographical place where the enumerated person usually resides implies a time element,” yet “the recommendations do not offer any time limit for considering oneself a usual resident of a place” (United Nations Statistics Division, 2004:3). Citing companion recommendations on the measurement of international migration statistics, the planning paper notes a trend toward defining “country of usual residence” as “the country in which he or she has a place to live where he or she normally spends the daily period of rest.” The planning paper offers no specific recommendation, but suggests that a revision of census principles for the 2010 round of censuses “could recommend time periods to be used for defining a ‘usual’ resident of a place” and offer “more guidance [to help determine] the treatment of persons who appear to have more than one residence” (United Nations Statistics Division, 2004:4). B.3 AUSTRALIA The Australian Bureau of Statistics (ABS) takes a multifaceted approach to residence concepts; the core quinquennial census is operated as a de facto census, enumerating people where they are on a designated Census Night.1 However, the census asks questions on usual residence and usual residence 5 years ago. Moreover, ABS formally defines “usual residence” in two ways, one for use in the census and the other for its regular demographic surveys. In 2001, as in previous censuses, the Australian census was conducted by field visits; enumerators dropped off forms at households and later returned 1 In 2001, that night was August 7; in 2006, it will be August 8.

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census to collect them. For the 2006 census, ABS plans to increase its reliance on mailback census responses (particularly for areas with high concentrations of apartment buildings), and it will permit Internet responses (Office for National Statistics, 2003). Though the census objective is a de facto count, information on usual residence is also gathered in order to compute the estimated resident population (ERP), ABS’ official population estimate series, which is used for electoral and fund allocation purposes. To convert from the census figures to ERP, three statistical adjustments are made: one each for estimated census coverage errors (undercount and overcount), Australian residents temporarily overseas, and backdating from Census Night to an official reference date of June 30. An ABS review of the “usual residence” concept concluded that it is “impractical…to devise a single standard concept of Usual Residence that satisfies the full range of analytical requirements of users and operational exigencies” (Australian Bureau of Statistics, 2004). In simplest terms, ABS considers a usual residence “the dwelling a person calls home and resides in on a permanent basis” and it is defined “based on the fact that each person has a basic attachment to a particular dwelling.” That attachment does not lend itself to a single definition, and so ABS defines two separate usual residence concepts: Attachment to the dwelling in which a person lives the majority of the time, which is the concept used in the quinquennial census. The formal definition allows for prospective definition of residence; the concept refers to the address at which “a person has lived or intends to live” for 6 months or more in a year, “even if they do not regard it as their home and do not have a strong social, economic or familial attachment to it.” Attachment to the dwelling which a person considers to be their “home,” or family home, which is used in the Australian Monthly Population Survey and other household surveys. This concept, based on self-perception, “embodies social, economic and familial attachment to a dwelling because it contains their household or family home.” Under this concept, a person may “be considered as a usual resident of the dwelling in which their family home is located even if they do not live there the majority of the time.” “As the majority of people live in their family home the majority of the time the same dwelling constitutes their usual residence in both cases.” Figure B-1 illustrates the form of the usual residence question as it is planned to be asked in the 2006 Australian census; as noted in the figure, the 2006 format is similar to the presentation used in 2001. ABS household surveys like the Monthly Population Survey depend greatly on enumerator interviews, including those by telephone. The ABS usual residence concept paper (Australian Bureau of Statistics, 2004) notes that the usual residence

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census NOTE: The 2006 version of the question is very similar to the one used in 2001; “Apartment/Flat/Unit number (if any)” has been added, and the 2001 instructions instructed “persons who now have no usual address” to write “no usual address” but did not specify where to write that text. The bulleted point on “persons who usually live in another country” was a new addition to the question in 2001. SOURCE: Trewin (2005). Figure B-1 Proposed form of basic usual residence questionnaire item (UR1), 2006 Census of Population and Housing, Australia question on the household survey forms first asks “What are the names of all the people who usually live here?” and then asks “Will any of these people be staying away tonight?” If a respondent answers yes to the latter question, he or she is then asked, “Will [name of person] be away for more than six weeks altogether?” If the answer to the third question is yes, that person’s record is retained for household (family) coding but they are excluded from being counted in the survey. The usual residence question has been on the Australian census form for every census since 1961, except for 1966. In 1961, only the person’s state or territory of usual residence was requested; in all other instances full address information was collected. A question asking respondents for their full address of usual residence 5 years ago has also appeared in all Australian censuses since 1971, and a question on usual residence 1 year ago became a fixture in 1976 (although there has been variation in whether a full address was requested or

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census . merely a state/territory). A planning document indicates that the three usual residence questions will be reviewed prior to the 2006 census and that “the standard for usual residence is under review” (Edwards, 2003:25), and similar language is included in the first report detailing content for the 2006 census (Trewin, 2005). B.4 CANADA The Canadian de jure census is conducted every 5 years. Since 1971, questionnaires have been dropped off by enumerators and are expected to be returned by mail. Broader mailout of questionnaires to two-thirds of census addresses was implemented in 2006, after completion of an address register (Office for National Statistics, 2003). The 2001 Canadian census form devotes a full page to a set of include/exclude instructions; this instruction block is reproduced in Figure B-2. The 32-page “Census Guide” prepared by Statistics Canada to provide additional information for respondents lists a basic rationale for the usual residence questions in Steps B and C: “These steps help you to decide who should be included and who should not be included in the questionnaire. They tell us that we have counted everyone we need to count and that we have not counted anyone twice.” For the 2006 Canadian census, the residence instructions are much the same as in 2001 but are edited for syntax, so that all the bulleted points read as complete phrases rather than colon-separated definitions. Some proposed changes are more substantial: The potentially confusing point on “ABSENT SPOUSES” is modified to read: “SPOUSES OR COMMON-LAW PARTNERS WHO LIVE ELSEWHERE while working or studying, but who return here periodically.” The point on persons in the country with some form of work permit is simplified to: “PERSONS FROM ANOTHER COUNTRY WITH A WORK OR STUDY PERMIT and family members living here with them.” Emphasis is added to parts of the last two bulleted points, on institutional residents and persons with no other home. They are now slated to read “PERSONS who usually live here but are now IN AN INSTITUTION (such as a home for the aged, a hospital or a prison), IF THEY HAVE BEEN THERE LESS THAN SIX MONTHS” and “PERSONS staying here on May 16, 2006, WHO HAVE NO USUAL HOME ELSEWHERE.”

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census Figure B-2 Residence instructions, 2001 Census of Population, Canada

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census Figure B-3 Basic residence questions, 2001 Census of Population, Canada If all the (similar to 2001) “do not include” instructions apply, respondents are still asked to fill in a circle but are prompted to give their name and telephone number before mailing in the (not completed) questionnaire. In 2001 the actual residence questions were presented as shown in Figure B-3; a coverage probe question asking the respondent to consider persons who might be missing is included in Step C. For 2006, Statistics Canada plans to keep the same listing of up to 10 household members, but the roster question will be preceded by a basic count query: “Including yourself, how many persons usually live here, at this address, as of May 16, 2006? Include all persons who usually live here, even if they are temporarily away.” [Because the new question begins with “including yourself,” the reminder in 2001’s Step B—“Don’t forget to include yourself!”—is omitted in 2006. The Step C coverage probe is identical to the 2001 version.

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census Though not a formal list of residence rules, Statistics Canada posted an overview of “2006 Census Collection” on the Internet that does illustrate some underlying residence concepts; see http://www222.statcan.ca/ccr01/ccr01_001_e.htm [8/1/06]. In particular, the document identifies some “special enumeration” categories: Students away from home are enumerated in both places but counted at the parental home. “Students attending school out of town but who return home when school is not in session should be included on their parents’ questionnaire, as part of the regular household.” However, because the “school residence is considered a collective dwelling,” they are required to complete the first two pages of the census form at the school. People living in “non-institutional collective units”—“inns, hotels, motels, campgrounds, YMCA/YWCAs, and military bases”—are enumerated there; if they are only “staying temporarily,” they are counted at their usual place of residence. Statistics Canada makes explicit that “residents at institutions such as detention facilities, hospitals, residences for senior citizens, orphanages or prisons are enumerated using the institution’s administrative records.” However, exception is made for “seniors who reside in institutions or residences with distinct, separate living quarters that do not blend with units such as those for chronic care;” if those persons are able to complete the census questionnaire, they are allowed to do so. As mentioned in the questionnaire instructions, they are counted at the facility if they have been there at least six months. The document also specifies that the objective of the census is to include “persons alive at midnight between May 15 and 16, 2006,” so that babies born on May 16 are not included. Though the document says that the census “will take place on Tuesday, May 16,” and that “on [that date], the majority of households will complete a census questionnaire and return it online or mail,” the document also includes a contradictory instruction: “householders are asked to complete the questionnaire…and return it either online or [by mail] by May 16th, Census Day.” The Canadian census includes a long-form sample with additional questions. A long-form question on usual residence 5 years ago has appeared in the Canadian quinquennial census since 1961; the question on usual residence 1 year ago first appeared in 1991. In 2001, respondents were asked for the city, province, and postal code (but not full street address) for the 1-year and 5-years-ago residences if they were located in Canada; if they were out of the country, only the name of the other country was requested. The 2006 versions of these questions appear identical to their 2001 counterparts.

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census B.5 ESTONIA Estonia’s most recent population census—and the first since reacquiring its independence—was conducted in 2000. The objective of the census was to characterize the population and housing characteristics of the nation as of the “census moment,” midnight (00:00) on March 31, 2000. The census was designed in accord with the UNECE (1998) guidelines. Section 7 of the Population and Housing Census Act of 1998, which authorized conduct of the census, mandated that the Estonian census include “persons who are in the Republic of Estonia at the moment of the Census [with exceptions noted below],” “persons who reside in the Republic of Estonia but who are in foreign states temporarily for a term of up to one year,” and “diplomatic staff of diplomatic missions and consular posts of the Republic of Estonia and their family members, who are in a foreign state at the moment of the Census.” The census explicitly excludes “diplomatic staff of foreign diplomatic missions and consular posts and their family members” as well as “persons in active service in a foreign army.” Though the focus is on a de facto count at the census moment, the Estonian census also collected de jure residence information, as well as usual residence in the previous (1989) census. Section 4 of the Census Rules for 2000 Population and Housing Census holds that “every enumerated person covered by the Census has to determine his/her permanent place of residence (permanent, main, usual place of residence), from the temporary absence of which he/she can elsewhere have a temporary place of residence. The permanent place of residence is the place of residence where he/she lives permanently, regardless of whether he/she is in-registered at that place or has the right to use the dwelling.” Consistent with the UNECE suggestions, the census rules hold that a person’s permanent or usual residence “is the place where he/she spends the majority of his/her daily night-rest.” If there is doubt as to which of two places should be judged the permanent residence, the rules say that “the place should be preferred where he/she intends to live for most of the time of the year.” According to the census rules, Estonia’s handling of certain special populations is quite similar to the current U.S. model, with some slight deviations. College students and “pupils of professional secondary or other educational institutions” are deemed to be usual residents of the place where they study and “not the place of residence of parents.” With regard to prisons and other institutions, “the permanent place of residence of persons who have lived in an institution for more than a year or who will stay there for more than a year is the institution.” However, persons in military service are counted at “the place where they departed from for military service.”

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census B.6 IRELAND Since 1951 the Central Statistical Office of Ireland has conducted a quinquennial census, with only slight deviations; the 1976 census was canceled due to budget constraints and replaced with an abbreviated census in 1979, and the most recent census in 2002 reflected a postponement from 2001 due to an outbreak of foot and mouth disease.2 The Irish census is oriented as a de facto count as of census night and instructs its respondents to fill out the census form on the designated night. The 2002 census of Ireland was conducted on Sunday, April 28, 2002. Page 1 of the census instrument includes the following instructions on “How to complete your Census form:” The form should be completed on the night of Sunday 28 April. Please answer questions about the household on page 2. Identify on page 3: all persons (including visitors) who spent the night of Sunday 28 April in the household; any household members who are normally resident in the household but who are temporarily away on the night of Sunday 28 April. Answer the questions beginning on page 4 for all persons present on the night of Sunday 28 April. Answer the questions on pages 22–23 in respect of any household members temporarily away on the night of Sunday 28 April. Sign the declaration on the back page. The draft form for the 2006 census, constructed following a 2004 pilot test, only modifies the date of the census. Page 3 of the 2002 questionnaire asks the respondent to “List every person who spent the night of Sunday 28 April in the household or who arrived the following morning not having been enumerated elsewhere.” A bulleted list provides further inclusion and exclusion instructions: INCLUDE all persons alive at midnight on Sunday 28 April. persons staying temporarily in the household. DO NOT INCLUDE babies born after midnight on Sunday 28 April. anyone who is temporarily away from home on the night of Sunday 28 April. However, these persons should be listed as being absent in List 2 below. 2 See http://www.cso.ie/census/When_is_Census.htm [5/30/05].

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census Students and pupils of regular schools as well as those attending special training schools or miscellaneous schools who were living in school dormitories, boarding houses or the like were enumerated at the places where they lived regardless of their period of stay. In-patients of hospitals were enumerated in the hospitals only when they had been hospitalized for 3 months or more at the census date. Otherwise, they were enumerated at their homes even if they were expected to be in the hospital for 3 months or more. Crews aboard ships except ships of the self-defence forces were enumerated at their residential places on land, if any. Otherwise, they were enumerated on the ships if the ships were of the Japanese flag and were at anchor at a port of Japan at the census date or if the ships left a port of Japan before the census date and entered a port of Japan within 5 days after the census date without calling at any foreign ports. Residents in the camps of the self-defence forces were enumerated in the camps. Crews aboard ships of the self-defence forces were enumerated at the places of the local general headquarters to which their ships belonged. Persons in prisons or detention houses whose penalties had been fixed and inmates of reformatories or women’s guidance homes were enumerated at those institutions. “In accordance with the rules described above, all persons living in Japan were enumerated whether they were foreigners or not.” However, foreign diplomats and military personnel and their families were excluded. The Japanese census asks for information on the length of stay at the usual residence as well as the usual residence 5 years ago. B.10 NEW ZEALAND New Zealand’s quinquennial census has traditionally been conducted through enumerator drop-off and pick-up of questionnaires; due to its de facto goal of recording the population present on census night, census forms were even distributed on overnight trains on the 2001 census night of March 6 (Office for National Statistics, 2003). The main product of the census is the “census night population count,” which includes all people who were “on New Zealand soil, on a vessel in New Zealand waters, or on a passage between New Zealand ports” on census night. This count includes “overseas residents and other people in diplomatic residences in New Zealand” as well as “foreign military personnel and their families located in New Zealand on census night” (Statistics New Zealand, 2001:11). Questionnaire probes also allow construction of the “census usually resident population count,” so that “if a person usually lives in Christchurch but was in Wellington on census night, they will be included in the census usually

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census resident population count for Christchurch and the census night population count for Wellington” (Statistics New Zealand, 2001:12). Like the Australian census, the New Zealand census is intended as a de facto count, but usual residence information is collected for purposes of generating estimates for fund allocation and other purposes. New Zealand’s official definition of “usual residence” is interesting because it is explicitly defined to the “meshblock” (equivalent to census blocks in the United States) level. The official definition used in the 2001 census is lengthy (Statistics New Zealand, 2001:16): Usual residence is the [census block] of the dwelling where a person considers himself or herself to usually reside, except in the following cases: people who board at another residence to attend primary or secondary school, and return to their parent’s(s’) or guardian’s(s’) home for the holidays, usually reside at the address of their parent(s) or guardian(s). Post-secondary students usually reside at the address where they live while studying children in joint custody usually reside at the place where they spend more nights, or if they spend equal amounts of time at each residence, they usually reside at the place where they are at the time of the census people who are in rest homes, hospitals, prisons or other institutions, usually reside where they consider themselves to live, and this may include the institution a person whose home is on any ship, boat or vessel permanently located in any harbour shall be deemed to usually reside at the wharf or landing place (or main wharf or landing place) of the harbour a person from another country who has lived, or intends to live, in New Zealand for 12 months or more usually resides at his or her address in New Zealand (as in external migration) people who spend equal amounts of time residing at different addresses, and can not decide which address is their usual residence, usually reside at the address they were at on census night, or if none of the above guidelines apply, the person usually resides at the address he or she was surveyed at. The definition of usual residence does not include a time-criterion and instead uses the approach of self definition. This is because a time criterion can lead to households and families being classified on an arbitrary basis. Furthermore, most people know where they usually live (reside) and as such this involves feelings of belonging, association and participation in and with a household. Address of usual residence has been a question in New Zealand censuses since 1921. A question asking how long the respondent had lived at that address was added in 1976, dropped in 1981, and has been asked in the 1986

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census and subsequent censuses. A question on usual residence 5 years ago has been asked since 1971; a question on usual residence 1 year ago appeared only in the 1971 and 1981 censuses (Statistics New Zealand, 2001:33). Despite the length of the formal Statistics New Zealand definition of usual residence, the census instruments for 2001 are relatively free of residence instructions or cues. However, the questionnaire includes repeated cues to the respondent to include babies in the enumeration. The New Zealand census relies on completion of two forms: a dwelling form, filled out by one member of the household that includes housing questions, and an individual form, one of which is completed for each person in the house. The principal residence instruction appears in a small column beside the first question on the dwelling form and reads: These people fill in an Individual Form here in this dwelling: everyone, including babies, who is spending the night of 6 March here; and anyone who arrives on 7 March, who has not filled in an Individual Form anywhere else. This includes babies. On the second page of the dwelling form, a household roster (up to 10 people) is constructed in question 4: “List all the people who are filling in a blue Individual Form here in this dwelling (and people having one filled in for them), starting with yourself as Person 1.” Two special instructions are highlighted in circles under person 1 (where relationship to household questions would normally appear): those read “Remember to list any babies who live here!” and “If a baby is aged under one year, print [graphic showing a mark of zero].” Several pages later in the dwelling form, question 18 asks, “Will everyone who usually lives in this dwelling fill in a blue Individual Form (or have one filled in for them here)?” If the answer is no, the respondent is prompted to go on to question 19, which asks, “How many people who usually live here WON’T fill in a blue Individual form here (and WON’T have one filled in for them here)?” Beside these two questions is a text box: Count as usually living here children away at boarding school people who are away on holiday, away for work, in hospital for a short time, etc. DON’T count university or other tertiary students who live somewhere else for most of the year For up to five such absentees, the dwelling form respondent is asked to provide the missing person’s name, age, gender, and relationship to respondent;

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census they are also asked whether the absentee is in New Zealand on census night and, if not, “how long altogether is s/he away from NZ?” (responses are “less than 12 months” or “12 months or more”). The blue-colored Individual Form is bilingual and contains only the instruction, “One of these forms must be filled in for every person in New Zealand on the night of 6 March 2001.” Question 5 asks, “Where do you usually live?” and a blank box is provided for respondents to write the full address. Question 6 asks how long the respondent has lived at that address (either “less than one year,” or a write-in for the number of years). Question 7 asks, “Where did you usually live 5 years ago on 6 March 1995?” and allows the responses “not born 5 years ago,” “at the address you gave in question 5,” “in New Zealand at another address” [blank for full address], and “NOT living in New Zealand” [blank to write in country name]. Question 8 asks for census night residence information: “On the night of Tuesday 6 March, what address are you at?” Responses are “at the address you gave in question 5,” “at another address” [blank for full address]. Based on feedback to initial content plans for the 2006 census, Statistics New Zealand reported that it was considering enhancing the geographic detail coded for absentees from their usual residence. In 2001, absentees were only coded to the meshblock (census block) of usual residence, not explicitly linked to specific residences. Final determination as to whether absentees would be repatriated back to their usual home of residence (for reconstruction of family statistics) was said to depend on final census tests (Statistics New Zealand, 2003). The standard of usual residence “was raised repeatedly during consultation with stakeholders,” with particular regard to “overseas students studying in New Zealand” and “New Zealand students who are away from home on census night.” The Statistics New Zealand planning document also acknowledges “known difficulties with this topic, such as a respondent’s interpretation of the word ‘usual’ ” (Statistics New Zealand, 2003:10). Though some analysts suggested reinstating a question on usual residence 1 year ago, Statistics New Zealand declined to include it in the 2006 census, judging that “usual residence five years ago gives the best indication of intercensal migration.” The bureau also declined to add questions on the number of moves made in the last year and the reasons for moving (Statistics New Zealand, 2003:11). B.11 SOUTH AFRICA The South African census of 2001 was intended to provide a de facto count of persons present in South Africa on the night of October 9–10, 2001. Limited information was collected on usual residence, which was based on where a person spends 4 or more nights a week. Question 11 asked “Does (the per-

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census Figure B-4 Questionnaire items to collect primary and secondary address information, 2000 Census of Population, Switzerland son) usually live in this household for at least four nights a week?” If no, respondent was asked a follow-up as question 11a: “Where does (this person) usually live?” If the usual residence is also the place of data collection (current residence), that could be indicated by checking a box; if different, respondents were supposed to answer with the province of residence and the name of the place (for South Africa) or country name (for residences in another country). The census form also asks if their place of usual residence is the same as it was 5 years ago, during the preceding census; if the answer is no, a follow-up prompts for the year in which the person moved (if more than one move was involved, the year of the most recent move was to be indicated). B.12 SWITZERLAND The Swiss de jure decennial census defines the “resident population” as “all persons who officially reside in Switzerland over a given period of time regardless of their citizenship, duration of residence and type of permit. Persons who do not officially reside in Switzerland (e.g. cross-border commuters working in Switzerland, tourists, visitors or business travellers) do not fall into this category” (http://www.bfs.admin.ch/bfs/portal/en/index/themen/volkszaehlung/definitionen.html [April 2005]). The 2000 Swiss census form collected address information on both a primary and a secondary residence with a very limited amount of space on the physical page of the personal questionnaire (see Figure B-4). Arguably, it may be too tight a space: the key follow-up question as to which of the two addresses is the place where “you mainly reside” is not well distinguished and could be missed or skipped.

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census Question 5 on the questionnaire is “Place of residence 5 years ago; where were you living on 5 December 1995?” (Because Switzerland’s censuses are decennial, the 5-year-ago request does not correspond with the immediately previous census.) Four responses are permitted: “At the same address as now (residence A),” “In the same commune (as residence A) but at another address,” “In another commune (specify),” or “Abroad” (respondent is asked for the country). B.13 UNITED KINGDOM The British decennial census was a de facto enumeration for most of its history, from 1801 to 1971. In 1981 and 1991, hybrid approaches collecting both de facto and de jure information were tried before the 2001 census switched solely to de jure counting (Smith, 2005). Early in the planning cycle leading to the 2011 United Kingdom census, the Office for National Statistics considered switching back from a de jure to a de facto count. They have since determined that the population base for the 2011 census will continue to be usual residents, augmented by visitors to the nation present on census night. The 2001 British decennial census was executed pursuant to the “Census Order 2000” passed by Parliament and enacted on March 15, 2000; the target census day for the 2001 count was April 29. Somewhat akin to the 31 formal residence rules for the 2000 U.S. census, the order formally defined “usual residence” with respect to a set of eight groups (see Table B-1). Specifically, the legislation defines “usually resident” as including: persons who have a usual address in England and Wales; are present at an address in England and Wales on census night and have no other usual address in England and Wales or elsewhere; in the case of Groups I and IV, are in full-time education and are residing at their term-time address; in the case of Groups II, III, VI and VII, have resided or intend to reside in the premises or vessel for a total period of 6 months or more beginning on or before census day; and in the case of Group V, are spending a period of 6 months or more in custody whether at the premises or elsewhere. A formal set of definitions published after the census provided an easier-to-parse definition of usual residence (Office for National Statistics, General Register Office for Scotland, and Northern Ireland Statistics and Research Agency, 2004:17): A usual resident is generally defined as someone who spends the majority of their time residing at that address. It includes

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census Table B-1 Usual Residence Categories as Delineated by the Census Order 2000, United Kingdom Group Premises, vessel orother place [of enumerations] Persons [to be counted] I Any dwelling or part of a dwelling (including a caravan, houseboat or other temporary building or structure used as living accommodation) occupied by a household (whether consisting of one or more persons). Every person who—(a) is usually resident in the dwelling or the part of the dwelling whether as a member of the household, paying guest or boarder, or as an employee of any such person; (b) not being a person to whom paragraph (a) above refers, is in full-time education and has a home address in the dwelling or the part of the dwelling. II Any hotel or guest house. Every person who is usually resident in the premises and who has not been included in any other return in the United Kingdom. III Any hospital, nursing home, religious or charitable community or other residential establishment whatsoever, not being an establishment mentioned elsewhere in this Schedule. Every person who is usually resident in the premises and who has not been included in any other return in the United Kingdom. IV Any residential school, college or other educational establishment. Every person who is usually resident in the premises. V Any civil prison or other place of detention. Every person who is usually resident in the premises and who has not been included in any other return in the United Kingdom VI Any vessel which is at a port in England and Wales at midnight ending census day, barracks, station or other premises under naval, military or air force discipline. Every person who is usually resident on the vessel or in the premises and who has not been included in any other return in the United Kingdom. VII Any vessel which is at a port in England and Wales at midnight ending census day not included in Group VI. Every person who is usually resident on the vessel and who has not been included in any other return in the United Kingdom. VIII Any other place not included in the above Groups. Every person who has no usual address.

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census people who usually live at that address but who are temporarily away from home (on holiday, visiting friends or relatives, or temporarily in a hospital or similar establishment) on Census Day; people who work away from home for part of the time, or who are members of the Armed Forces; a baby born before 30th April 2001 even if it was still in hospital; and people present on Census Day, even if temporarily, who have no other usual address…. The usual resident population did not include: people present at an address on Census Day whose usual address was elsewhere; or people away from their home address who had been living, or intended to live, in a special establishment such as a residential home, nursing home or hospital for six months or more (they were enumerated as usually resident at the special establishment). College and boarding school students were to be counted at their school address. British military personnel permanently stationed in Northern Ireland were to be recorded at their “actual address of residence unless they were married and unaccompanied by the spouse,” in which case they were to be counted at the address shared with the spouse. The 2001 count represented a shift toward a more pure de jure census; while the 1991 census sought information from both usual residents and visitors present on census night, the 2001 instrument collected data only on usual residents. Space to complete rosters of household members and visitors took up a full second page of the individual census form used in the United Kingdom census of 2001. First, respondents are asked to complete Table 1 pursuant to the following instructions: List all members of your household who usually live at this address, including yourself. Start with the Householder or Joint Householders. Include anyone who is temporarily away from home on the night of 29 April 2001 who usually lives at this address. Include schoolchildren and students if they live at this address during the school, college or university term. Also include schoolchildren and students who are away from home during the school, college or university term and for whom only basic information is required. Include any baby born before 30 April 2001, even if still in hospital. Include people with more than one address if they live at this address for the majority of time.

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census Include anyone who is staying with you who has no other usual address. Remember to include a spouse or partner who works away from home, or is a member of the armed forces, and usually lives at this address. Space is then provided for the names of 10 household members. Directly below Table 1, Table 2 asks the respondent to list up to 5 names and addresses for visitors to the household: “To help you complete the form you may use Table 2 to list any visitors at this address, on the night of 29 April 2001, who usually live elsewhere.” The instructions further direct that “if there are only visitors at this address,” only the first five questions on the nature of the housing unit should be completed; “no further questions need to be answered.” Later in the questionnaire, at the beginning of collection of personal characteristics, full-time students living elsewhere are immediately screened from further questioning. If a person answers “yes” to question 5, “Are you a schoolchild or student in full-time education?,” they are asked “Do you live at the address shown on the front of this form during the school, college or university term?” If the answer is yes, the student is guided through the complete set of person-level questions (race, employment, health, etc.). If the answer is “no,” they are routed to the end of questions for that person. The possible switch back to a de facto or “persons present” count in 2011 was described in the initial design document (Office for National Statistics, 2004). The rationale for the switch included the argument that the usual residence definition in 2001 “introduced ambiguity about who should be included and may have resulted in higher non-response amongst certain population groups” (Office for National Statistics, 2004:3). The design document also noted strong arguments from the census user base on the need for counts of the daytime or working population of areas, for better planning of services. Further consideration reported by Smith (2005) raised serious issues of discontinuity if the de facto model is adopted. Specifically, 2011 results would not be directly comparable with the 2001 census returns, but they could also differ from post-2011 counts, for which the Office for National Statistics would likely rely heavily on administrative records data for which usual residence is coded. Maintaining a count of the usual resident base but with additional emphasis on collecting some information from visitors (particularly “workers who are not usually resident but contributing to the national economy” and other short-term visitors) was judged to be the best compromise and recommended by Smith (2005). This recommendation was corroborated by Office for National Statistics (2005) and Stokes (2005b), and issued as a formal Office for National Statistics recommendation in Stokes (2005a).

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census Stokes (2005a) further notes that exact definitions for both “usual resident” and “visitor” remain to be decided. The Office for National Statistics is still considering the addition of a series of questions related to second residence: “whether a person lives at another address for part of the year, and if so, what this address is, what the address is used for, and the amount of time spent at this address” (Courtney, 2005a:3). The switch is due to perceived increases in the complexity of living situations in the United Kingdom; analysis of auxiliary data sources including the stand-alone Survey of English Housing suggests a fluctuating but generally increasing trend in ownership of second residences (Courtney, 2005a,b).

OCR for page 303
Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census This page intentionally left blank.