Inner TRIM3 Masthead
TRIM3 Logo

Imputing Information About Non-Citizens (AlienPrep) version 13.8

Version History


The AlienPrep module imputes variables needed to simulate the eligibility of immigrants for SSI, AFDC/TANF, SNAP benefits, and Medicaid. These person-level variables are:
  1. YearEnteredUS - The year in which the person entered the U.S. to stay.
  2. YearsLivingInUS - The number of years an immigrant has lived in the U.S.
  3. PreReformImmigrant - Did the legal alien or refugee enter this country on or before 22 August 1996?
  4. VeteransExemption - Does the immigrant qualify for a veteran’s exemption?
  5. YearsFor40Q - The number of years in which the immigrant has worked in this country. This applies only to persons age 15 or older.
  6. FortyQuartersExemption - Does the person qualify for the forty-quarters-of-work exemption? Children may qualify for this exemption as well as adults, depending on the number of quarters worked by their parents.
  7. SponsorDeemingExempt - Is this person exempt from sponsor deeming?
  8. The Unit - Identifying Groups of Immigrants

1. Year Entered US

YearEnteredUS is computed from the CPS ASEC categorical variable YearInUS, which specifies a narrow range of years when a person came to the U.S. to live. This variable was made available in CPS ASEC data beginning with the March 1994 survey and is updated annually. Beginning with the conversion of the March 2015 survey data for use by TRIM3, EditedYearInUS was also made available. Except for very few cases that are edited for the immigrant imputation, it is identical to YearInUS. AlienPrep program rule YearInUSVar may be used by analysts to specify either YearInUS or EditedYearInUS for AlienPrep simulations. Specific year ranges for YearInUS codes are shown in the TRIM3 dictionary.

Year ranges are defined in the array program rule YearBrackets. There is an array entry for each range, and the maximum year for each range is stored in YearBrackets. A four-digit value for YearEnteredUS is imputed for each unit with YearInUSVar>0. For non-immigrant units for which YearInUSVar is zero, YearEnteredUS is also zero. For immigrants who entered the country before 1950 (YearInUSVar = 1), YearEnteredUS is set to 1949.

To impute YearEnteredUS (for persons with YearInUSVar > 1), AlienPrep performs the following steps.

  1. Calculate the birthyear for each person:
    • Birthyear = Data year - Age
    • For example, if the simulation year is 2014 (March 2015 CPS), a person who is 5 years old will have a birthyear of 2014-5 = 2009. In reality, 5 year olds born in January-March would have been born in 2010--but this method of calculation holds true for the majority of persons.
  2. For each unit:
    • Determine the range of years (FirstYear to LastYear) in which the unit could have entered the U.S. For most units, FirstYear is the first year in the range indicated by YearInUSVar, and LastYear is the last year in the range indicated by YearInUSVar. But if a person’s birthyear comes after the first year in the range indicated by YearInUSVar, then the first year that the person could have come to the US is his or her birthyear. Since we are assuming all of the people in each unit came to the U.S. at the same time, FirstYear will be determined by the age of the youngest member of the unit.
For example, consider a family consisting of a man, his wife, their 16 year old child, and the man’s brother. The man, wife, and child have YearInUSVar=6, indicating they entered the country in 1975-1979. The man’s brother has a YearInUSVar=10, indicating he entered the country in 1986-1987.

If we are using the 1993 input data (CPS 1994), then the child’s birthyear is 1993-16 = 1977. The man, wife, and child reported entering the country in 1975-1979, but since the child was not born until 1977, they must have entered between 1977 and 1979. So for this unit, FirstYear will be 1977, and LastYear will be 1979. The man’s 30 year old brother entered in 1986-1987. His FirstYear will be 1986 and his LastYear will be 1987.

If there is anyone in the unit whose birth year falls after the last year in the YearInUSVar range, an error message is generated.

Imputing a Single Year in Which the Unit Arrived.

Once FirstYear and LastYear have been defined, a year is randomly assigned to the unit using probabililties stored in the array program rule YearProbabilities for years prior to 2009. Probabilities stored in array program rule YearProbabilities2 are used for year 2009 and later. There is an array entry for each year. The first element in the YearProbabilities array is for 1949, the second for 1950, and so on.

AlienPrep assigns YearEnteredUS as follows:
  1. Calculate TotalProbability --the sum of the probabilities for each year beginning with FirstYear and ending with LastYear.
  2. Generate a random number, RandomNumYearEnteredUS, for the unit.
  3. Calculate CumulativeProbability = probability for FirstYear / TotalProbability. If the random number is less than or equal to CumulativeProbability, set YearEnteredUS = FirstYear. Otherwise increase CumulativeProbability by (probability for the next year/TotalProbability). If the random number is less than or equal to CumulativeProbability, set YearEnteredUS equal to that year. Continue in this way until a YearEnteredUS has been assigned.
  4. Store the resulting YearEnteredUS on the record of each person in the group.
In the example above, if the values entered into the YearProbabilities program rule for 1977, 1978, and 1979 were .3, .2, and .3 respectively, then
TotalProbability = .3 + .2 + .3 = .8

The initial cumulative probability would be .3/.8 = .375. If the random number is less than or equal to that, say .233, then the module would assign YearEnteredUS=1977. But if the random number is greater than that, it would increase CumulativeProbability as follows:
CumulativeProbability=CumulativeProbability+.2/.8= .375+.25=.625.

Let’s say that the random number is .8. If so, then increasing CumulativeProbability for the final year (.625 + .3/.8) brings us to 1.00, and we would assign YearEnteredUS = 1979.

Special Imputations for Recent Immigrants

Legal aliens arriving after August 22, 1996 were denied eligibility for most benefits, so a legal alien arriving between 1996 and 1998 who reports receiving benefits is likely to have entered the country on or before August 22, 1996. However, there are exceptions to this rule. Certain states provide benefits to legal aliens arriving after that date. There are four program rules which indicate whether receipt of benefits will be ignored for persons in a particular state for the purpose of determining year of entry-- IgnoreReportedSSI, IgnoreReportedTANF, IgnoreReportedFoodStamps, and IgnoreReportedMedicaid.

For each member of the unit who is a legal alien (CorrectedCitizenship=1), AlienPrep determines for the following conditions. If any of the following conditions hold true for any legal alien in the group, then the module sets YearEnteredUS for the group to 1996.

  • IgnoreReportedSSI{state}=0 and SSI >0 and AllocFlagSSIReceipt=0
  • IgnoreReportedTANF{state}=0 and AFDC>0 and AllocFlagAmountofPublicAssist=0 and AllocFlagPublicAssistanceType=0
  • IgnoreReportedFoodStamps{state}=0 and HasFoodStampRecipients=1 and AllocFlagFoodStamps=0 (Note, these are household variables so it is possible that other members of the household received the SNAP benefits, but we won’t worry about this detail)
  • IgnoreReportedMedicaid{state}=0 and HealthMedicaidCoverage=1 and AllocFlagMedicaidCoverage=0

2. YearsLivingInUS

The number of years in which the person has lived in the U.S. is calculated as follows:

YearsLivingInUS = Data year - YearEnteredUS

Note that the data year is one year less than the survey year, e.g., the data year for the 2015 ASEC survey is 2014. If YearEnteredUS is zero, YearsLivingInUS is defined to be zero as well. Also, some people may be imputed to arrive in the survey year (Data year + 1). For these people, YearsLivingInUS will be zero.

3. PreReformImmigrant

If a person is not a legal alien or refugee (CorrectedCitizenship is not 1 or 3), PreReformImmigrant is defined as 0. The calculation of PreReformImmigrant for legal aliens and refugees depends on the year in which the entered the U.S. (YearEnteredUS):

YearEnteredUS PreReformImmigrant
1995 or earlier 1
1996 If any legal alien in the immigrant’s unit reported receiving public assistance benefits (see "Special Imputations for Recent Immigrants" in section on YearEnteredUS), PreReformImmigrant is 1.

Otherwise, pick a random number, RandomNumPreReformImmigrant, for the unit. If this number is lower than ProbabilityPreReform, then PreReformImmigrant is 1. If it is greater than ProbabilityPreReform, PreReformImmigrant is 0.
1997 or later 0
4. VeteransExemption This section creates two new person-level output variables: ExemptVeteran indicates whether the person is a legal alien or refugee who is also a veteran or current member of the armed forces. ExemptVeteranDependent identifies people who are legal aliens or refugees, and who are also either the spouse or unmarried dependent child of a veteran or current armed forces member. Unlike the variables created in the previous sections, different people in the same unit will have different values for these.

ExemptVeteran is determined as follows:
Value Conditions
1
  • The person is a legal alien or alien refugee (CorrectedCitizenship=1 or 3).
  • The person is a veteran (VeteranStatus is between 1 and 5) or a current member of the armed forces (PopulationStatus=2 or LastMajorOccupation=14).
0 Otherwise.
If ExemptVeteran is zero, ExemptVeteranDependent is determined as follows:
Value Conditions
1
  • The person is a legal alien or alien refugee (CorrectedCitizenship=1 or 3).
  • The person is either the spouse or the dependent child (See below) of a person who is a veteran or current armed forces member (See above) and who is either a legal alien, naturalized citizen, alien refugee, naturalized refugee, or native (CorrectedCitizenship is 1,2,3,4, or 7).
0 Otherwise.
A person is considered a child under the following conditions:
  • The person is not married (MaritalStatus is not between 2 and 5).
  • The person is not the head of a household (HhFamilyRelation is not 1 or 32).
  • The person is under 18 (Age < 18) or under 22 and attending school at least one month of the year (MonthlyLaborForceActivity = 7).

5. YearsFor40Q

This section and the FortyQuartersExemption section provide specifications for imputing whether a legal immigrant qualifies for the 40 quarters of work exemption. If a legal immigrant has acquired 40 quarters of work in the United States, then he or she is exempt from rules denying aliens eligibility for assistance under state and federal government programs. In determining eligibility for the 40 Quarters of Work Exemption, the immigrant can count his or her own quarters of work, those worked by a spouse during the time that the couple are married, and those worked by the immigrant’s parents before the child turned 18. In order for a quarter of work to count, it must be work that is creditable for Social Security benefits. Therefore, we will assume that work performed by an illegal alien or an alien only temporarily in the country does not count. In addition, we will ignore work performed before a person has finished his or her education (or reached age 22, whichever comes first). The imputation was developed using data from the 1993 Survey of Income and Program Participation (SIPP). The design of the imputation is very much driven by data limitations in the SIPP. Section VII.1.1 defines variables created in support of the imputations. Section VII.1.2 describes the imputation of years of work.

For simulation year 1998 and future years, we will probably want to take reported receipt of benefits under the various transfer programs into account. Pre-reform immigrants already participating in these programs were allowed to continue receiving benefits through much of 1997, so reported receipt of benefits does not provide any useful information for the 40 quarters imputation for simulation year 1997.

Definitions of Intermediate Variables Used in the Imputation

For purposes of describing the methodology used in the imputation, it is helpful to first define certain variables used in the methodology. Some of these variables are actually available as output under the category "For Debugging", while others exist only for the duration of the simulation.

CitizenStatus
Our goal is to determine whether persons with CitizenStatus=2 are eligible for the 40 quarter exemption. To do this, we’ll need to impute the person’s own years of work, as well as the years of work of certain other family members who have CitizenStatus=1 or CitizenStatus=2. Work performed by persons with CitizenStatus=3 should not be counted toward anyone’s 40 quarter exemption.

1 = naturalized or native-born citizen (CorrectedCitizenship=2, 4, or 7)
2 = legal alien or alien refugee (CorrectedCitizenship=1 or 3)
3 = non-immigrants and illegal aliens (CorrectedCitizenship=5 or 6)
PercentOfPoverty
Usually when dealing with poverty in TRIM3, we perform the calculations ourselves rather than relying on the CPS supplied variables. However, for the purpose of AlienPrep, it is fine to use PovertyRate (which is derived during the conversion process from CPS variables).

For persons in related subfamilies (FamilyType=3), use the PovertyRate from the primary family in the household (the family with FamilyType=1), rather than the PovertyRate for the related subfamily. The reason this is done is that the PovertyRate for the related subfamily is based just on the subfamily’s income and family characteristics, whereas the PovertyRate stored with the primary family is based on the income and family characteristics of the entire family. The latter is the definition to use here.

AlienPrep
Uses three different break-downs for PercentOfPoverty:
  1. <100%, 100%-<200% and 200%+ (PovertyRate: 1-3, 4-7, and 8+)
  2. <200% and 200%+ (PovertyRate: 1-7 and 8+)
  3. For summary tables only: <150%, 150%+ (PovertyRate: 1-5, 6+)
CurrentWorker
Is a person considered a "current worker" for the purpose of the imputation?

This definition of CurrentWorker is designed to match, as closely as possible, the definition in the SIPP data upon which the imputation is based. Treat a person as a CurrentWorker if MonthlyWeeksWorked is greater than zero in any or all of the months: December, January, February, and March.

0 = not considered a current worker
1 = considered a current worker, i.e. worked at least one week in December, Jan, Feb, or March
PotentialWorkingYears
The years a person could have worked

Estimate the age at which the person could have begun doing "real" work, based on the person’s current age and the amount of education he or she has completed. First, assign WorkingAge based on the chart below. HighestGradeCompleted is a TRIM3 Variable, with values as indicated in the table below.

Education
Completed
Corresponding
HighestGradeCompleted Values
Assign
WorkingAge Of
Ninth Grade or Less 0-10 15
Tenth Grade 11 16
Eleventh or Twelfth Grade 12-13 18
AA degree or College but no Degree 14-15 20
Bachelor’s degree or higher 17-18 22

Figure out the person’s total number of potential working years. Don’t allow a value below zero.

PotentialWorkingYears=max(Age-WorkingAge,0)

Cap PotentialWorkingYears at the number of years living in the United States (as imputed by AlienPrep).

PotentialWorkingYears=min(PotentialWorkingYears,YearsLivingInUS)

If the person is a full-time student (InSchoolFullOrPartTime = 1), PotentialWorkingYears are 0.

Impute Years of Qualifying Work for 40 Quarter Exemption (YearsFor40Q).

We’ll need to impute years of qualifying work for each person age 15+ with CitizenStatus=2 (legal alien or alien refugee). For legal alien and alien refugee children, and for legal alien and alien refugee adults with fewer than 10 years of work, we will have to impute years of work for certain additional family members (as discussed in the FortyQuartersExemption section).

YearsFor40Q should be a real value (fractions of years are allowed). If the imputation results in a value greater than 10, YearsFor40Q should be set to 10. (YearsFor40Q is topcoded at 10, because "10" is the magic number for our purposes here, and the imputations aren’t accurate enough to show the exact number of years greater than 10). However, for use in computing FortyQuarterExemption and for optional microdata output, we produce YearsFor40QUncapped, the uncapped version of YearsFor40Q.

  1. If a person has no working-age years (PotentialWorkingYears = 0) in the United States, set YearsFor40Q = 0:
  2. Otherwise, if the person receives Social Security Benefits (SocialSecurityOrRrr > 0) and PotentialWorkingYears >=10, then assign YearsFor40Q=10
  3. Otherwise, if the person has 1 to 18 potential working years in US,
  4. perform the following imputation for YearsFor40Q:
    • Impute the percentage of the person's PotentialWorkingYears that have been worked, and convert this percentage to a number of years. The imputation relies on program rules showing the probabilities that a person worked: none, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90% or 100% of his or her potential working years. The actual program rules vary by citizen status and gender. For current workers (CurrentWorker = 1) the probabilities vary by percent of poverty. Program rules for current workers are shown in Table 1. For persons not currently working, the probabilities also vary by PotentialWorkingYears (and, in certain cases, by percent of poverty). Program rules for persons not currently working are shown in table 2.

Table 1

40 Quarter Imputation Program Rules:

Current Workers with 18 or fewer PotentialWorkingYears

Probability A Person Works A Given Percentage of Potential Working Years (0-100)

Citizen
Status
Gender Percent of Poverty Program Rule
1 M <200% PctWorkingMaleCitUnd200
1 M 200%+ PctWorkingMaleCit200Plus
2 M <100% PctWorkingMaleAlienUnd100
2 M 100-<200% PctWorkingMaleAlien100to199
2 M 200%+ PctWorkingMaleAlien200Plus
1 F <200% PctWorkingFemCitUnd200
1 F 200%+ PctWorkingFemCit200Plus
2 F <100% PctWorkingFemAlienUnd100
2 F 100-<200% PctWorkingFemAlien100to199
2 F 200%+ PctWorkingFemAlien200Plus

Table 2

40 Quarter Imputation Program Rules:

Persons not currently working, with 18 or fewer PotentialWorkingYears

Probability Person Works Given Percent of Potential Working Years (0-100)

Citizen Status Gender Potential Working
Years
Percent of Poverty Program Rule
1 or 2 M 1-8 all PctNotWorkMaleUnd9
1 or 2 M 9-18 all PctNotWorkMale9To18
1 F 1-8 all PctNotWorkFemCitUnd9
1 F 9-18 all PctNotWorkFemCit9To18
2 F 1-8 <200% PctNotWorkFemAlienUnd9A
2 F 1-8 200%+ PctNotWorkFemAlienUnd9B
2 F 9-18 <200% PctNotWorkFemAlien9To18A
2 F 9-18 200%+ PctNotWorkFemAlien9To18B


Steps for performing the imputation:

    • Obtain the set of probabilities corresponding to the person’s current working status, citizen status, gender, potential working years, and percent of poverty.
    • Generate a random number for use with this imputation, RandomNumYearsFor40Q.
    • Set CumulativeProbability = probability of working 0% of his or her PotentialWorkingYears. If the random number is less than or equal to CumulativeProbability, then set YearsFor40Q=0. Otherwise increase CumulativeProbability by the probability the person worked 10% of his or her PotentialWorkingYears. If the random number is less than or equal to CumulativeProbability, then set YearsFor40Q=(.10 * PotentialWorkingYears). Don’t round YearsFor40Q to an integer—keep as a real value. If the random number is greater than CumulativeProbability, increase CumulativeProbability by the probability the person worked 20% of his or her PotentialWorkingYears. If the random number is less than or equal to CumulativeProbability, then set YearsFor40Q=(.20*PotentialWorkingYears). Otherwise, continue in this fashion until YearsFor40Q has been assigned
    • The probabilities in the program rule will sum to 1.00 (give or take 0.01 - due to rounding). If the random number is greater than the final CumulativeProbability (as would be the case if the final CumulativeProbability was 0.991 and the random number was 0.998) assign YearsFor40Q=PotentialWorkingYears.

    If the person is currently working and a value of 0 has been assigned, set YearsFor40Q = 0.5 (we’ll give token credit for work performed during the year). Cap YearsFor40Q at 10 (as mentioned before, save the uncapped value under the name YearsFor40QUncapped).
4) Otherwise, if the person has 19 (PotentialWorkingYears >= 19) or more potential working years in US, perform the following imputation for YearsFor40Q:
    The imputation is similar to that for persons with 1-18 PotentialWorkingYears, except that each program rule gives the probability that a particular number of years has been worked (0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10+). Therefore, there is no need to multiply the result by PotentialWorkingYears. As before, assign 0.5 years of work to current workers imputed to work 0 years.

    Table 3 shows the program rules for current workers, and Table 4 shows the program rules for persons not currently working.

Table 3

40 Quarter Imputation Program Rules:

Current Workers with 19 or more PotentialWorkingYears

Probability Person Works Given Number of Years (0-10)

Citizen
Status
Gender Percent of Poverty Program Rule
1 M <200% YrsWorkingMaleCitUnd200
1 M 200%+ YrsWorkingMaleCit200Plus
2 M <200% YrsWorkingMaleAlienUnd200
2 M 200%+ YrsWorkingMaleAlien200Plus
1 F <200% YrsWorkingFemCitUnd200
1 F 200%+ YrsWorkingFemCit200Plus
2 F <200% YrsWorkingFemAlienUnd200
2 F 200%+ YrsWorkingFemAlien200Plus

Table 4

40 Quarter Imputation Program Rules:

Persons not currently working, with 19 or more PotentialWorkingYears

Probability Person Works Given Number of Years (0-10)

Citizen Status Gender Potential Working
Years
Percent of Poverty Program Rule
1 or 2 M 19+ all YrsNotWorkMale
1 F 19+ all YrsNotWorkFemCit
2 F 19+ all YrsNotWorkFemAlien

6. FortyQuartersExemption

The forty-quarters exemption is calculated only for legal aliens and refugees (CitizenStatus = 2). For all other people, it is 0.

Initialize FortyQuartersExemption to 0.

    If YearsFor40Q is equal to 10 for a given person, the person qualifies for the 40 quarters exemption. Set FortyQuartersExemption to 1.

    Otherwise, if the person has a spouse present (SpousesLineNumber is not 0), add the spouse’s YearsFor40Q to the person’s own YearsFor40Q. If the result is greater than or equal to 10, set FortyQuartersExemption to 2. (Reminder, YearsFor40Q should be 0 for all persons with CitizenStatus=3—so if the spouse has CitizenStatus=3, there will be nothing to add here.)

    Otherwise, if this is a child under the age of 18 (HhFamilyRelation is 3-7, 13, 15 or 27), count the parents’ YearsFor40Q, as well as the child’s (and the child’s spouse, if any), to determine if the child is covered by the FortyQuartersExemption. If the total years is at least 10, then set FortyQuartersExemption to 3. Note that if the child is the head of a sub-family, the parents will be the people in the FamilyWithSubfamilies who are either the head or spouse (FamilyReferenceRelation is 1 or 2) of the main family (FamilyType = 1). In all other cases, the parents are simply the head and spouse of the immediate family.

    Otherwise, if this is a person age 18 or older who has a parent or parents living in the household (HhFamilyRelation is 9-12), the parents’ years of work before the child turned 18 count toward the FortyQuartersExemption. (Note, this includes work performed before the child’s birth and before the child came to the US). We’ll approximate this as follows:

    1. For each parent, calculate Years40QBeforeChild18---the years parent worked toward the 40 quarter exemption before child turned 18.

        (a)Calculate number of potential working years of parent before child turned 18:

          WorkYearsBeforeKid18=

          PotentialWorkingYears{Parent} - (Age{Child} -17)

          If WorkYearsBeforeKid18 <= 0, then Years40QBeforeChild18= 0 -- this is all you need to do for this parent. Otherwise....

        (b)Calculate number of potential working years before child turned 18 as a percentage of parent’s total potential working years

          PercentYearsBeforeKid18=

          WorkYearsBeforeKid18/PotentialWorkingYears{Parent}

        (c)Calculate Years40QBeforeChild18:

          (1) If the parent has 18 or fewer potential working years, then we have an estimate of the years worked--up to and including PotentialWorkingYears (YearsFor40QUncapped). Assume years of work have been distributed evenly over all PotentialWorkingYears, and assign years worked before child turned 18 accordingly.

            Years40QBeforeChild18 = PercentYearsBeforeKid18*YearsFor40QUncapped{parent}

          (2) Make the same assumption if the parent has more than 18 potential working years, but has worked fewer than 10 of these. (There is no "uncapped" value for parents with more than 18 potential working years, so use YearsFor40Q instead of YearsFor40QUncapped).

            Years40QBeforeChild18 = PercentYearsBeforeKid18*YearsFor40Q{parent}

          (3) If the parent has more than 18 potential working years, and has worked 10 or more of these, we don’t actually know how many years have been worked. Our simplifying assumption will be that the parent worked each of the potential working years before the child turned 18 (obviously an overestimate, but is closer to the truth than assuming the parent worked none of these years, and we don’t have an alternative method).

            Years40QBeforeChild18=

            Min(WorkYearsBeforeKid18,YearsFor40Q{parent})

    2. If YearsFor40Q{Child}+YearsFor40Q{Child’s spouse) + Years40QBeforeChild18{parent1}+ Years40QBeforeChild18{parent2} >= 10 then set FortyQuartersExemption=3


    Otherwise, if this person is a legal alien (as opposed to a refugee) who has been in the U.S. for more than 5 years, and does not qualify for an veteran exemption (either as a veteran or a dependent), the person may be imputed to qualify for the 40-quarters exemption based on the reported receipt of certain benefits as follows:

    1. If the person is a true SSI reporter (i.e. the reporting of SSI was not allocated to the person by Census), set FortyQuartersExemption to 4.
    2. If the person is a true Medicaid reporter (i.e. the reporting of Medicaid was not allocated to the person by Census) and the rule IgnoreReportedMedicaidFor40Qtrs = 0, set FortyQuartersExemption to 5.


    Note that imputation based on reported receipt of benefits applies regardless of whether the person is married, is under 18, or is 18 or older and living with parents.

7. SponsorDeemingExempt

Legal immigrants who are not refugees may be ineligible for benefits because the income of their sponsor is deemed available for their use. SponsorDeemingExempt identifies immigrants who should be considered exempt from this sponsor deeming test because they reported receiving SSI, AFDC, or SNAP benefits within a certain number of years, as defined in the program rules DeemingYearsSSI, DeemingYearsAFDC, and DeemingYearsFoodStamps. Each of these state-level program rules contains separate values for pre-reform and post-reform immigrants. If AllocFlag665 indicates the entire supplement was allocated, then no one in the unit is considered to receive public benefits.

For pre-reform immigrants (PreReformImmigrant=1), if anyone in an immigrant’s family reports SSI receipt not imputed by Census and has been in the U.S. (YearsLivingInUS) for the number of years in the pre-reform value of DeemingYearsSSI or less, then all members of that family are considered exempt sfrom sponsor deeming. If DeemingYearsSSI is 0, reported receipt of SSI does not exempt the family from sponsor deeming.

For post-reform immigrants (PreReformImmigrant=0), if anyone in the immigrant’s family reports SSI receipt not imputed by Census and has been in the U.S. (YearsLivingInUS) for the number of years in the post-reform value of DeemingYearsSSI or less, then all members of that family are considered exempt from sponsor deeming. If DeemingYearsSSI is 0, reported receipt of SSI does not exempt the family from sponsor deeming.

Tests for AFDC, SNAP, CHIP and Medicaid are similar to the above, with the program rules and input variables changed appropriately. For the exact input variables used to determine receipt of SSI, AFDC, or SNAP benefits, see Special Imputations for Recent Immigrants above.

8. The Unit

For the YearEnteredUS, YearsLivingInUS, and PreReformImmmigrant imputations, we will want to assign the same value to all members of the FamiliesWithSubfamilies family who have identical values for YearInUSVar. Each such group is considered a unit for AlienPrep. Units can be easily identified using the output variable UnitID.

Rule AdditionalDeemingExemptRate indicates the rate at which Legal Permanent Residents who are not identified as exempt from sponsor deeming based on public benefit receipt, and who are not identified as having a likely family sponsor within their household, are randomly chosen as having no family sponsor (SponsorDeemingExempt=1).

If AllocFlag665 indicates the entire supplement was allocated, then no one in the unit is considered to receive public benefits.