CPS Basic Notes
Information available
CPS Basic Questions
Year of Entry
When did <NAME> come to live in the United States?
Variable: INUSYR
Same as on the ACS
Country of Birth
In what country was <NAME> you born?
Variable: NATVTY
Citizenship
Is <NAME> a citizen of the United States?
Was <NAME> born a citizen of the United States?
Did <NAME> become a citizen of the United States through naturalization?
Variable: CITIZN
Same codes as on the ACS
All three variables are available on the input file for 2nd and later interviews.
All three are asked in the first interview, or if the answer is “Don’t know.”
In the first interview, the three questions are asked with the CPS Basic questions. Supplement questions get asked after all of those.
The residence one year ago section is NOT available on the CPS Basic, only the ASEC supplement.
Survey Practice
Most people in the household will have their information answered by proxy.
The CPS is given only via telephone (CATI) or in person (CAPI). NO mailout.
Time limit: average less than 10 minutes added per household, averaged across all households
Sampling
Sample size: The unweighted N of FB (not born a citizen) in the CPS is around 14,000, N native is around 122,000. The N of people not born in the 50 states + DC is around 16,000.
Universe: We can make Armed Forces persons eligible for the Supplement.
Sampling:
Regarding the possibility of oversampling FB and Hispanic households: there’s a strong concern about nonresponse due to oversampling for multiple surveys, i.e., for both the ASEC (Nov) and the ATUS (2 months, select 1 for each). This is a note from Note from Greg Weyland in the CPS branch of DSD.
Aug 2008-MIS 1 & 5 will be rotations 4 & 8 in November 2008, among which all Hispanic will be eligible for ASEC:MIS1: 2009,2010; MIS 5: 2008,2009
Aug 2008-MIS 8 IF EITHER (a)ALL HH members non-Hispanic white only and with kids <19 OR (b) Non-Hispanic racial minorities with or without children will get American Time Use Survey in Nov, Dec and ASEC Feb 2009. This adds about 3000 cases.
For people who are not U.S. citizens.
Anticipated N: 7,000-9,000
Anticipated N people naturalized since first CPS interview: 150-200 [based on OIS 2006 naturalization number of 702,589]
Currently, citizenship is asked only on the first interview of the CPS Basic questionnaire and is not updated in subsequent CPS interviews. However, for noncitizens the status could change between the first interview and the current interview. By asking for changes in citizenship for noncitizens, the Citizenship section can observe changes in naturalization status since the first interview. The resulting data could be used to measure (i) the frequency of change and (ii) correspondence with naturalization rates from administrative data such as from the Office of Immigration Statistics (OIS). Also, the data can be used to assess the effect of changes in naturalization on interpretation of the relationship between citizenship and outcome characteristics.
Question
Source: one of the three CPS basic questions used to create variable CITIZN
Asked only for noncitizens, whose naturalization status may have changed since the first interview. We use the “Is <NAME> a citizen…” question rather than the “naturalization” question from the CPS Basic. This is because due to the 1994 cognitive test, this question is less sensitive and more easily understood, especially by native born proxy respondents. Through the universe restriction, we will be able to infer that the person is a naturalized citizen. This assumes that citizens have not changed their citizenship since the first interview.
Wellens, Tracy. 1994. The Cognitive Evaluation of the Nativity Questions for the Current Population Survey. U.S. Census Bureau paper prepared for presentation at the annual meeting of the American Association for Public Opinion Research, Danvers, MA, May 1994.
Anticipated N: 15,000 – 16,000
This section has a number of purposes. First, the CPS Basic question on the year came to “live” in the U.S. will be re-asked, in second and later interviews, in order to examine the quality and consistency of the current year of entry question between multiple reports, over time and potentially different report sources- self or proxy. Additionally, this section will be used to assess the validity of alternate measures of the two main uses of year of entry information: (1) year of entry and (2) total time spent in the U.S.
Measures of year of entry are
Year came to “live” in the U.S. (YOE) [current measure]
First year of entry
Most recent year of entry.
Measures of total time spent in the U.S. will be calculated as
TS1. (Survey year) – (YOE Year came to “live” in the U.S.) [current measure]
TS2. (Survey year) – (First year entered)
TS3. (Survey year) – (First year entered) – (Total time spent outside of the U.S. since First year)
TS4. (Survey year) – (Most recent entry year)
Harris, P.; R. Bhaskar, C. Shook-Finucane, and L. Ericson. 2007. Evaluation Report Covering Place of Birth, U.S. Citizenship Status, and Year of Arrival. U.S. Census Bureau.
Kerwin, J.; S. Heltemes, H. Berry, D. Nelson, and M. Popovic. 2005. Cognitive Testing of Proposed Items on International Migration for the American Community Survey. Rockville, MD: Westat, for the U.S. Census Bureau.
Poros, M. and A. Orum. 2005. Year of Arrival Second Cognitive Test Contract- Report on Cognitive Testing of Proposed International Migration Questions for the American Community Survey. U.S. Census Bureau: ISS contract.
Redstone, I. and D. Massey. 2004. Coming to Stay: An Analysis of the U.S. Census Question on Immigrants’ Year of Arrival. Demography, 41(4), 721-738.
Source: INUSY2 -CPS Basic (INUSYR)
This question will be asked in order to assess the amount of variability that is due to survey methods- time and proxy reporting- for the current CPS (and ACS) measure of year came to “live” in the U.S. (CPS Basic variable INUSYR). Currently, the question is only asked in the first CPS interview, if it is a replacement household, or, in 2nd or later interviews, if the information is “Don’t know.” By asking for verification of the original year of entry question, we can assess the proportion of variability that is due to time (recall) and also to response source (proxy or self-response). We can also study the effect of multiple entry scenarios by checking inconsistent came to “live” years with the first and most recent years of entry.
These questions will be used to validate the current year of entry question, which measures the year came to “live” in the U.S. Also, data on multiple years of entry can be compared to OIS data on new visas and immigrants. INUSB4 and FRSTYR will be used to check for prior entries to the year came to “live” in the U.S. Information on most recent year of entry will be gathered from YRLEFT and LASTYR (see below).
In answering these questions about alternate years of entry, respondents will be asked to consider only stays 2 or more months inside or outside of the U.S. The 2-month rule is employed because it corresponds to the ACS residence rules, as well as a heuristic definition of what is a “brief” trip or visit (the wording used in cognitive tests of 2004). The term “2 months or more” is used because it may be easier for respondents with English as a second language than the term “at least 2 months.”
Source: New
This question is used to facilitate skip patterns. If there have been no exits from the U.S., then no more information is needed for this section, for this person. It is also to help the respondent start thinking about travel and personal migration event history. Travel is meant to include all trips of any kind, of any duration, for any purpose (including business, weddings, funerals), to anywhere (including Canada, Mexico, "home", or elsewhere). It also includes "living" outside the U.S. for any period of time. This would be an especially important concept for people who came to the U.S. prior to the year they came to "live" in the U.S. and returned home or elsewhere to "live" during the interim period.
This question, along with TRV2MO, replaced YOENUM, and is a much easier cognitive task, especially since it is a yes/no question and does not restrict exits to durations longer than 2 months.
Source: New
This is an alternate measure of total or accumulated time of U.S. exposure that will account for multiple exits, time not physically in the U.S. since the first year they came to the U.S. It will be compared to (i) the currently used measure (survey year minus year came to “live” in the U.S.) and (ii) another measure to be created in the supplement (survey year minus first year of entry FRSTYR). In answering the question, respondents are asked to consider ALL exits, not just those that are 2 months or longer. This is intended to account for migrants, for instance circular migrants, who make frequent stays or exits under 2 months’ duration and, therefore, spend a substantial proportion of the year outside of the U.S. While the data may be somewhat imprecise, there is currently no information of this type at the Census Bureau. This question will be important as a comparison to and assessment of the traditional measure.
The question (TIMOUT) is worded with time outside the U.S. since people may be more likely to remember and add total amounts of time for travel outside the U.S. than intervals inside the U.S. between trips abroad. TIMOUT is offered with the options "Less than one year" and "One year or more." This is to reduce the difficulty of adding up the duration of multiple trips outside the U.S. Many people who have taken many multiple short trips will be able to answer “Less than a year” without having to think longer on it. Precision down to months or weeks is unnecessary.
The question is placed right after the question on any exits from the U.S., since the exits may be of any duration. Later, exits are limited to trips of two months or longer as it was found to be cognitively easier to impose restrictions later in the survey. Also, for information on the total number of years (YRSOUT), this question may be subject to recall and calculation issues as well as proxy source imprecision. Cognitive testing found respondents had little difficulty understanding question in the second round. Also, most were able to give a reasonable estimate for TIMOUT (more or less than a year). For those who were asked for YRSOUT (total number of years outside the U.S.), a few people responded with a range. While all of these latter respondents were people who have been in the U.S. for over 20 years, two-thirds of respondents who have been in the U.S. for over 20 years did respond with a precise answer. In response to BLS’s concerns about recall and precision, we have restricted the universe on this question to people who report a length of residence in the U.S. of 15 or fewer years, i.e., they entered in 1993 or later. After data collection, the responses will be analyzed for prevalence and order of magnitude before being categorized and topcoded meaningfully.
TRV2MO
Source: New
This question, along with ANYOUT, replaces YOENUM in former versions. This is an easy (Yes/No) question to facilitate skips into and out of questions about most recent year of entry with a 2 month stay. Asking this question is usually easier for the respondent than adding the 2 month condition to the direct year of entry question.
This may be redundant for people who named a FRSTYR different from INUSY2, but instead of making the text more complicated or routing these people to another questions (complicating skip patterns), Anna Chan (SRD) suggested we just ask the same question for everyone, for simplicity’s sake.
It is asked after TIMOUT/ YRSOUT total time/years outside the U.S. because it is cognitively easier to restrict (duration) later in the interview than to restrict then open up (e.g. to trips of all duration).
Respondents will be asked to consider only stays 2 or more months outside of the U.S. The 2-month rule is employed because it corresponds to the ACS residence rules, as well as a heuristic definition of what is a “brief” trip or visit (the wording used in cognitive tests of 2004).
YRLEFT / OUTYRLAST
Source: New
These questions will be used to check if there is a later entry than the year came to “live,” but only if the individual has exited and reentered the U.S. since the first year of entry. Otherwise, the most recent year of entry will be assumed to be the same as the year came to “live” in the U.S. Due to the skip pattern from TRV2MO, this implies that the most recent year of exit and re-entry refer to an exit of duration at least 2 months. These two questions YRLEFT-LASTYR are meant to approximate asking the start and end years for the most recent “spell” of time outside the U.S. that is 2 months or more. During preliminary analyses of the cognitive testing results, SRD recommended this approach, but a Blaise-based instrument cannot input two fields on the same screen. Thus, using these two questions is the final recommendation.
There are multiple reasons for using a “spell” type of approach for most recent year of entry information. In Cognitive Testing Phase I, many respondents wanted to provide the most recent year they left instead of the year they returned. Asking for year of exit helps respondents walk through this step on the way to getting the year of most recent entry, rather than asking a single, direct question on most recent year of entry. Also, most respondents returned in the same year they left.
In phase II, the question sequence tested was: most recent year left for a period of 2 months or more (YRLEFT), whether or not they returned the same year (YOELASTYR), and, if not, what year did the person return (LASTYR). For most people, this new sequence worked, was understood and the information provided was as the question intended. However, the sequence did not work in the situation of someone experiencing consecutive periods outside the U.S., with intermittent brief trips to the U.S., as in the case of the Armed Forces. While the sequence was intended to get the MOST recent year of entry to the U.S., the question asked: "Did [you/NAME] come back to the United States that same year?" Respondents in this situation would correctly answer the sequence in an unintended way by providing the year of the intermittent brief trip back to the U.S. rather than the final, most recent year of entry. Even though the estimated number of such cases is small at 32 (0.2% of Census 2000 foreign born were in the Armed Forces), SRD thought it would be cleaner overall to use a “spell” approach.
For YRLEFT, an invalid response category is offered: “Not applicable, Never outside the U.S. for at least 2 months.” In theory, this response should not happen, since all relevant cases should already have been routed from UCKY2 to UCKY3. Just in case, we have this option.
For LASTYR, we offer the response “Currently outside the United States.” In the Phase 1 testing, there was one case where a household member was still temporarily away. This can happen on the CPS, due to the residency concept of "usual residence" rather than "current residence" (as on the ACS). These people will also be listed on the list of “Emigrants and Residents Abroad” and will also be asked intended year of return in YOELAST. There is second option for people currently outside the U.S.: “Does not plan to return to the U.S.” Theoretically, no one should need this response, since anyone getting this question should already be considered a household member, i.e., is a ‘usual resident,’ and not 'moved out.' However, the rostering and membership rules are quite complex, with more detailed updating for interviews MIS 1 and 5 only. Thus, there may be some slippage and this response is offered for these cases.
RESIDENCE ONE YEAR AGO SECTION
For all people 1 year or older.
Anticipated N: 136,000-138,000
Anticipated N who lived outside the US one year ago: 500-700 [based on .4% of cases in CPS ASEC 2006]
While detailed information on residence one year ago is regularly collected on the CPS ASEC Supplement, only selected questions are asked on the Migration Supplement. The questions on living in the same housing unit (MIGSAM, NXTSAM) and on country of residence one year ago (MIGCN, NXTCN) will be collected in the CPS ASEC Supplement format, but detailed information on one year ago locale within the U.S. will not be collected on the Migration Supplement.
These data will be used to estimate native and foreign-born 1-year international in-migration over the same time period as the emigration information collected in the Residents and Emigrants Abroad section. Additionally, this information can be another check with the year of entry variations from the year of entry section, especially for those who arrived in the last one or two years. Finally, data on living at the same address one year ago (MIGSAM, NXTSAM) will serve the practical purpose of an eligibility screen for the Residents and Emigrants Abroad Section and also as an independent variable in analyses. Such analyses would include the characteristics of households screened into or out of the Residents and Emigrants Abroad Section and substantive analyses on mobility patterns, jointly considering internal and international migration.
Questions
MIGSAM, NXTSAM
MIGCN, NXTCN
MIGALL
MIGM
Source: CPS ASEC One-year migration questions MIGSAM, NXTSAM, MIGALL MIGM, MIGCN1, NXTCN1
This section is modeled after the ASEC, which collects information in a time saving manner that begins with the reference person (MIGSAM, MIGCN), then all other household members at once (MIGALL). In other words, the reference person’s (not necessarily respondent’s) residence one year ago information is obtained first (MIGSAM, MIGCN). Then all other household members are screened for whether or not they ALL lived with the reference person one year ago (not necessarily in the same housing unit) (MIGALL). For household members who lived with the reference person one year ago, residence abroad one year ago is then automatically coded to be the same as the reference person’s. For household members who did not live with the reference person one year ago, residence one year ago information is asked for each individual (NXTSAM, NXTCN).
RESIDENTS AND EMIGRANTS ABROAD SECTION
For all households in which the proxy respondent lived at that address one year ago.
Anticipated N, households: 46,000-48,500 [based on 87.9% of people in CPS ASEC 2006 lived at the same address, and 55,128 Households in CPS Aug 2006]
Anticipated N, 1-year out-migrants: 300-600 foreign born, 125-300 native
[foreign born numbers 2-4%, based on J. Van Hook and F. Bean (2007) 4.7% CI: (4.2,4.7). Native numbers 0.10-0.25%, based on 262,921 living on select military bases in September 2002 (Schachter IDSEM contract)]
The Residents and Emigrants Abroad section will provide information about emigrants and others living outside the U.S., a topic for which there is very little contemporary information. There is a strong need for data on international out-migration, since the U.S. Census Bureau is mandated to produce annual population estimates, which are affected by the out-migration component. Although the method proposed here is survey-based and has a smaller sample size than other data sources, it will provide a national-level estimate of emigration that will account for emigrants and residents abroad who may not be observed in other emigration estimation techniques. For instance, members of the U.S. Armed Forces often are not observed in other nations’ censuses, so that emigration estimates based on international census data will undercount such persons. Furthermore, survey data can provide information on emigrants’ characteristics that may be omitted from international censuses. The 1-year emigration estimates created from these data will also be used with data from the Residence One Year Ago section to create estimates of net international migration and population projections. In addition to estimates of migration components, the data collected in this section will be used in substantive analyses to examine the characteristics of emigrants and others living outside the U.S. for future work on estimating emigration.
A version of the Residents and Emigrants Abroad section was collected as the “Emigration and Americans Living Abroad Supplement” to the CPS in June 1991, November 1989, June 1988, and July 1987 (See Woodrow-Lafield, 1996; Woodrow, 1990). The Migration supplement proposed for August 2008 is a reinstatement of those supplements, but there are a number of differences. First, the scope of the supplement is expanded, so that the supplement overall includes topics other than just emigration and emigrants. Second, the scope of the information regarding emigrants and residents abroad is more focused on demographic characteristics highly relevant to the construction of the U.S. Census Bureau’s annual population estimates (sex, age, citizenship, country of birth) and a few items to help understand the population abroad (relationship to household reference person, country of residence, current main activity and U.S. Armed Forces or government employment affiliation).
Someone in the household must have lived in this unit one year ago (MIGSAM or NXTSAM=1) in order to have proxy information on former residents. The questions, however, will be still be asked of the respondent, even if she or he did not live in the unit one year ago. This is because of the great difficulty and frequent attempts to get an interview from a specific person in the household in the case that the respondent did not live at that address one year ago. Once a household is screened into the Residents and Emigrants Abroad section, other screening questions (HHEMIG, UREMIG) are asked regarding former household members, those who lived at that address one year ago but who now live outside the U.S. The sampling and estimation method proposed here is based on all households and relationships (UREMIG, HHEMIG), including nonrelatives. It is also a one-year estimate that is unlikely to be substantially affected by mortality. In contrast, the sampling approach for the emigration supplements from 1987 to 1991 was based on a multiplicity approach that was dependent on surviving close relatives.
The sampling method for this section has a number of advantages compared to the previous emigration supplements. The housing unit sampling here has little risk of bias due to multiple counting. The screening question refers to a specific residence at a specific time point, “one year ago.” Since it is unlikely that people have more than one usual residence at a single time point, this greatly reduces the likelihood that an emigrant could be twice counted. In contrast, the multiplicity sampling method is network based, counting non-deceased close relatives (parents, siblings, and children) who ever lived in the U.S. at any time point before returning abroad. The multiplicity method increases the likelihood of multiple counting for each family member, whereas the sampling method proposed here is household-based and unlikely to overestimate emigration.
Another advantage is that this method counts some emigrants who do not have any close relatives living in the U.S., while the multiplicity method only counts closely related individuals. Thus, while this method is known to be an underestimate, since it cannot observe when entire households move outside the U.S., it is less of an underestimate than under the multiplicity sampling method. The U.S. Census Bureau’s Population Division has already consulted the Demographic Surveys Division (DSD) and the Demographic Statistical Methods Division (DSMD) on sampling and weighting issues, as well as the Central Statistics Office of the Republic of Ireland and the Office for National Statistics in the U.K., which have a similar sampling approach to that proposed here.
Finally, this method uses a one-year reference period, which matches the reference period for international in-migration on the CPS ASEC supplement and the Residents and Emigrants Abroad section. This one-year reference has an additional advantage that mortality is expected to have little impact on the resulting estimates. In contrast, the multiplicity method, which refers to currently living close relatives who resided in the U.S. at any time point, is subject to the potentially large and uncertain mortality rates of both U.S. resident relatives and those of emigrants. Utilizing the relatively short one-year reference time minimizes the effect of mortality and also clarifies which mortality rates to use (U.S.-based), if adjustments are needed
The regular household weights can be used for each emigrant or resident living abroad. However, post-collection analyses may show that it is necessary to exclude cases in which the respondent did not live in the unit one year ago, but someone else in the household did live at that address one year ago. If so, then the regular household weight may need an adjustment for noninterview or high “don’t know” or refusal rates.
An alternative to the regular household weight would be to match an emigrant’s demographic characteristics (age, sex, country of birth, citizenship) to those of a current resident, and apply the person weight of the current resident to the emigrant. It is uncertain whether or not this would be appropriate, since (1) the characteristics and population controls used to adjust residents’ person weight (race, Hispanic origin) are not exactly same as those available for emigrants’ (country of birth, citizenship), and (2) some of the conditions adjusted for, such as nonresponse and undercoverage, may apply only to the survey process for U.S. residents and not to emigrants. Overall, there may not be much difference in using a matched person weight versus the regular household weight.
Tom Moore of Demographic Statistical Methods Division’s (DSMD) Supplements Branch and Gregg Robinson of Population Division were both consulted regarding appropriate weighting for the emigrants sampled in this supplement.
Also, Helen Hughes from the U.K.’s Office for National Statistics and Helen Cahill of the Republic of Ireland’s Central Statistics Office were both consulted regarding adjusting the estimates to account for the unobserved emigration of entire households. Both countries have similar emigration questions to the HHEMIG question proposed in this supplement. The U.K. had too few cases to make any adjustments. The Republic of Ireland uses a figure 6%, which is “borne of experience” (personal correspondence).
Mulder, T.; B. Guzman and A. Brittingham. 2001. Evaluating Components of International Migration: Foreign-Born Emigrants. U.S. Census Bureau: Population Division Working Paper No., 62.
Woodrow, K. 1990. Estimating Unknown Immigration Flows with the Current Population Survey. Paper presented at the annual meeting of the American Association of Geographers held in Toronto, Canada.
Woodrow-Lafield, K. 1996. Emigration from the USA: Multiplicity survey evidence. Population Research and Policy Review, 15, 171-199.
UREMIG
WHOOUT
Source: New
These questions are intended to make sure any current household members living abroad actually get counted on the residents abroad section. UREMIG allows the respondent to define “living” here and “living” outside the U.S. There is no two-month rule on stays either inside or outside the U.S. For more information, see the explanation below, under HHEMIG.
During Phase I testing, it became apparent that sometimes people on the household membership roster are currently living abroad. On the CPS Basic, changes to the roster are only probed (MAWAY) in interviews 1 and 5. So, UREMIG & WHOOUT are intended to make sure all household members who are living abroad in order are added to EMIGALL, the list of people now living abroad. Some members of the household may not have lived in the housing unit one year ago, but in order to simplify the number of skip pattern questions, we decided to ask this section for all household members currently living abroad. With the Residence One Year Ago information, we can later exclude people who were not living in the housing unit one year ago. The number of people getting routed into this section via this question will be very small. These people also get a reduced number of emigration questions since, as household members, they have been asked demographic information already.
Regarding the universe, in the case that someone is added to EMIGALL from the Year of Entry section AND the respondent still says “NO” the person is not living abroad: We will keep the person on EMIGALL and ask questions from this section on that person. This is likely a very small number of cases, and the number of questions is truncated anyway so it will not add much time to the survey. Also, if we are concerned about consistency, we can later exclude these cases from emigration estimates, and also do analyses about the main activity of these people who have been outside the U.S. for 2+ months but not considered “living” abroad.
Source: New
This is the second screen to check for non-household-members who are emigrants or residents living abroad, i.e., nonresident persons who (1) lived in the unit one year ago and (2) are now living abroad. The “one year ago” wording was used rather than a specific date, such as “week of Aug 12th, 2007”, in order to match the one-year migration questions on the CPS ASEC.
The wording is consistent with the Residence One Year Ago section in allowing the respondent to define “living” outside the U.S. and “living” in this house one year ago. While this is somewhat ambiguous, this is consistent with the RESOYA section, which we cannot alter, due to its many years’ presence on the CPS and ACS. Also, the alternative is very confusing and wordy, with the 2-month requirements in the U.S. and abroad. Third, this language may have more face validity for measuring whether or not someone is an ‘emigrant.’ The main concern with the language "living" is that the sample size is expected to be smaller without an objective time-based definition (2-month stays). However, the face validity using “living” may be better for measuring the concept of emigration.
There is no definition counting stays in the U.S. or abroad of two months (including intended 2 month stays). Discussion with SRD concluded with the 2-month requirements and definitions would be burdensome and confusing to the respondent. A two-month stay is used in the Year of Entry section partially to correspond to the ACS residency rule, but also because it is heuristically a cutoff point for a (temporary) migration, rather than a brief trip or visit. However, in this section, no comparisons can be made with the ACS. Estimates of both in- and out-migration would be constructed from the CPS supplement alone, using the Residence One Year ago section and the Residents and Emigrants Abroad Section. Thus, it is necessary to work with CPS Basic household membership definition of usual residence, which includes as household members people who are temporarily away, such as students, armed forces members, and people on business. For a full description of the residence rules, see the FR interviewing manual. Note: ACS residence rules ask for current residence and a two-month stay at that current residence.
Source: New (note: CPS basic household roster asks for both first and last names)
If there are any emigrants, get a list of names for reference purposes for asking the Residents and Emigrants Abroad questions. Also, for households in rotations 5 through 8, we could also cross check the data from the supplement’s list of emigrants and their characteristics to the CPS basic roster and person characteristics from the first interview 12-15 months prior.
People on the EMIALL list who are current household members should skip EMISEX through EMINAT below, since they have already been asked about in the Demographic section of the CPS.
Source:
EMISEX- CPS Basic, reworded to ask if “male or female” rather than for emigrant’s “sex”
EMIAGE- CPS 1988 Emigration Supplement, reworded to ask simply “how old” is emigrant rather than “”How old was this person on his/her last birthday?”
Emigrant’s demographic characteristics- sex, age- to be used in construction of one-year emigration estimates, a component of the net international migration (NIM) estimates produced annually in Estimates and Projections Area.
Source: CPS Basic question wording, with mainly Census 2000 response categories (collapsed boarders, roomers, etc. with all nonrelatives, additional categories for other relatives)
Emigrant’s relationship to the household reference person. This information is useful for cross-validation when matching to household rosters from 12-15 months ago. It is also useful substantively, as a measure of international ties and an indicator of potential intent to return. The relationship to reference person, as opposed to the respondent, is used to match the CPS Basic question on household members’ relationship to reference person.
Source: same as CPS Basic (NATVTY)
Emigrant’s country of birth to be used in construction of one-year emigration estimates, a component of the net international migration (NIM) estimates produced annually in Estimates and Projections Area. Also used to determine routing patterns for citizenship questions, similar to CPS Basic.
Source:
EMICIT: CPS Basic question wording (CITIZN), but with the ACS response categories that apply to this universe (not born in the 50 states + DC, Puerto Rico, or other U.S. island areas)
EMINAT: CPS Basic question CITPB
Emigrant’s citizenship status, also to be used in construction of one-year emigration estimates. Also of substantive interest as a factor in potential return migration to the U.S. Multiple questions are used here in order to accommodate DK/RF (Don’t know/ Refuse) responses on EMICOB and EMINAT, in which case we can still get an “unspecified U.S. citizen” information from EMICIT.
Source: 1988 June CPS Emigration Supplement, slightly different word order (In what country is <emig1> living now?) than the original (“In what country does this person now live?”)
Emigrant’s current country of residence. This information could be used to compare the results of this supplement’s estimates with other methods in progress in ISS that utilize information from other countries’ censuses and surveys.
People on the EMIALL who are current household members should skip to this question, since they have already been asked about in the Demographic section of the CPS.
Source: 1988 June CPS Emigration Supplement for all questions. EMIACT response category 2 changed from “Keeping house” to current CPS Basic occupation wording “Taking care of house or family”
Emigrant’s status with the U.S. government- as a member of or dependent of a member of the U.S. Armed Forces, or as an employee or dependent of an employee of the U.S. Government- and the emigrant’s activity otherwise. The four questions on status with the U.S. Government distinguish between persons who are very likely temporarily residing abroad and all others. The question on main activity (EMIACT) is of substantive interest, particularly as a potential indicator of temporary (school) or permanent (retired) emigration. All together, these questions could be used to infer an emigrant’s intent to return and could thus be informative for estimates and projections of net international migration.
TRANSFERS SECTION
All households.
Anticipated N, households with ANY transfers: 1000-1700 [based on New Immigrant Survey, foreign born LPRs: 15% gave to a spouse in the last 12 months, 15% received from a spouse, 15% gave to an adult child, 11% received from an adult child. For all-native households: a guess of 1%, for Armed Forces, students, second generation, etc.]
This section is intended to obtain basic information about household transfers and remittances across international borders, both giving and receiving. This is a relatively simple approach compared to the potentially complex nature of transfers. However, it will provide basic information that has never been obtained by the U.S. Census Bureau, and the information at this level is still of interest to outside organizations such as the Bureau of Economic Analysis, the United Nations, the World Bank, and the Inter-American Development Bank. Moreover, the resulting data have the advantage of being nationally representative and comprehensive with respect to including information on (1) both native and foreign-born households, and on (2) giving and receiving transfers.
Baluja, K. 2005. Migration Survey Research (MSR) Deliverable 3- Review of existing surveys administered by non-federal organizations in the U.S. U.S. Census Bureau: ISS contract.
Schachter, J. and K. Baluja. 2006. Migration Survey Research (MSR) Deliverable 4 and International Data Sources and Emigration 2 (IDSEM2) Deliverable 2- Review of a sample of existing surveys administered by federal organizations in the U.S. U.S. Census Bureau: ISS contract.
Schachter, J. 2005. International Data Sources and Emigration (IDSEM) Deliverable 4- Migration Module. U.S. Census Bureau: ISS contract.
Schachter, J. 2005. International Data Sources and Emigration (IDSEM) Deliverable 2- Data Inventory of International Migration Statistics. U.S. Census Bureau: ISS contract.
Questions
GIVANY / TRSEND
GIVNUM / TRSENDNUM
GIVAMT / TRSENDAMT
RECANY / TRREC
RECNUM / TRRECNUM
RECAMT / TRRECAMT
Source: New, but similar to questions from the New Immigrant Survey, Survey of Remittance Senders: US to Latin America. Also consulted census information for Guam and St. Lucia, and suggested migration module by Jason Schachter on IDSEM contract Deliverable 4.1.
These questions ascertain if there are household transfers across international borders in both directions, give to or receive from, within the last year, the number of remittances in each direction, and the total amount in each direction. This information is for substantive analyses, to provide a general idea of both the prevalence of and magnitude of transfers, and also as a measure of international ties. Total number of transfers is asked instead of regularity (e.g. weekly, monthly) since individual household members may transfer to and from many other individuals at different frequencies and amounts. Transfers TO someone else is worded with “Given or Sent” since people may transfer money by giving it to someone to give in person or through other means, rather than literally “sending” the money.
To ease respondents’ burden and recall ability for answering for the entire household, the reference time period is one year. Also, transfers of interest are limited to monetary transfers to and from relatives or friends, i.e., personal transfers, and do not include transfers to or from organizations. Since the questions are asked at the household level, the answers may be somewhat imprecise, but the intent of these questions is to gain information on the order of magnitude of transfers. The respondent is asked to answer with a specific number of times or precise dollar amounts, but the resulting data will be categorized after collection. There will likely be much variation in the frequency and amount of transfers per household so that the total dollar amount will likely have a large range of responses. This large, unknown range makes it difficult to ascertain a priori the number of and cutoff points for categories to use. Thus, the data will be categorized after data collection and analyses.
The universe includes all transfers by anyone in the household, even if the person is no longer a current member of the household. The transfers information of interest is for the last 12 months, not ‘current’ transfers behavior. During round II, there was a case of someone giving money to someone with the awareness that the money would be used to purchase items for a religious organization. Thus, we further clarified the definition of transfers to include all person-to-person transfers, and no person-to-organization, organization-to-person, or organization-to-organization transfers. ISS branch chief Elizabeth Grieco, who has an extensive background studying remittances, was key in all decisions regarding definitions of transfers for this survey.
Also during cognitive testing round II, the question of whether or not loans should be included arose, and we originally decided to include these as transfers, even if they may not be gifts. However, after discussion with Michael Mann and others from the Bureau of Economic Analysis (BEA), ISS decided to exclude loans from the definition of transfers in order to make the resulting data more useful to the BEA. They considered this issue a “top priority” since loans are not included in the definition of Balance of Payments by the United Nations Statistical Division (UNSD) Technical Sub-group on Movement of Natural Persons (TSG). Since the current version of the Migration Supplement cannot add questions to distinguish loans from other transfers, Estimates and Projects Area (E and P) and ISS decided to exclude loans from the definition of monetary transfers.
Finally, a comment from BLS suggested the use of the term “gifts of money” instead of “money gifts.” ISS has incorporated this suggestion into the survey questionnaire.
| File Type | application/msword |
| File Title | Notes on the Jun 4, 2007 4pm version of the CPS Aug 2008 Supplement |
| Author | chiu0303 |
| Last Modified By | griec001 |
| File Modified | 2008-06-03 |
| File Created | 2008-06-03 |