NLSY79 Appendix 29: Date of Interview Current Status Variables

National Longitudinal Survey of Youth - 1979 Cohort

NLSY79 Appendix 29: Date of Interview Current Status Variables

The recently released 1979-2016 NLSY79 online database contains a set of variables for survey years 1980-2016 with information on the respondent’s labor force status at each date of interview. The 2016 variables are:

T57716.00    DOI_EMPLOYED    DATE OF INTERVIEW STATUS – EMPLOYED
T57717.00   DOI_HOURS_WORKED   DATE OF INTERVIEW STATUS – HOURS WORKED
T57718.00   DOI_RETIRED   DATE OF INTERVIEW STATUS – RETIRED
T57719.00   DOI_DISABLED    DATE OF INTERVIEW STATUS – DISABLED

Below is a description of these variables and the process by which they were constructed. Similar variables are being constructed for survey year 1979 and will be released at a future date.

Variable Description

In survey years 1979-1998 and 2006, the NLSY79 included a heavily used variable called Employment Status Recode (ESR) for each year. These variables represented the respondent's employment status during the survey week and were based on a very specific set of questions that mirrored the actual monthly Current Population Survey. Because ESR can only be computed from those specific questions, it cannot be computed in a strictly equivalent way in survey years in which the CPS section was not fielded as part of the NLSY79.

Many inquiries have been received concerning the missing ESR variables in non-CPS survey years. The Date of Interview Current Status (DOI) variables are intended to help provide a similar concise picture of a respondent's current status on the date of interview. The DOI variables provide a snapshot, capturing whether on the interview date the respondent:

  • was employed (DOI_EMPLOYED)
  • worked more or less than 30 hours per week (DOI_HOURS_WORKED)
  • considered herself/himself retired or reported current retirement status during the course of the survey, regardless of employment status (DOI_RETIRED)
  • considered herself/himself disabled, or reported current disability status in the course of the survey, regardless of employment status (DOI_DISABLED)

Users should note that the working/retired/disabled statuses are not necessarily mutually exclusive within single survey years, or continuous through multiple survey years. For example, a respondent may report that s/he is retired and disabled, but also working in some capacity at a single interview point. Likewise, a respondent who reports herself/himself disabled at one point in time may not be disabled (or provides no evidence of disability) at a later interview date. Similarly, someone who reports retirement at an earlier survey date may be working and not be identified as retired at a later survey date. For instance, respondents who have reported retirement from the military or a civil service/law enforcement position at a relatively early age may go on to a second full-time career and be identified only as employed in later years. An examination of current status variables in multiple years is necessary to determine if a respondent ever identified as retired/disabled, even if their status variables in the most current survey years do not reflect these reports.

Users should note that because the Date of Interview Current Status variables are based on weekly arrays, users may encounter a subset of cases that are assigned an "employed" status in these variables but report that they are "not currently working" in the actual survey.

Variable Creation

DOI_EMPLOYED – DATE OF INTERVIEW STATUS – EMPLOYED
These variables are based on codes in the Work History WEEKLY LABOR STATUS array. If a job number or a code for active military enlistment is present in the array for the week of interview, the respondent is assigned a current status of "1 - working." If the array for that week contains a code for unemployed or out of the labor force, the respondent is assigned a current status of "0 - not working."

DOI_HOURS_WORKED – DATE OF INTERVIEW STATUS – HOURS WORKED
The hours worked are determined from the Work History HOURS WORKED array. Those with 0 hours worked are assigned a -4 code. Respondents who worked between 1-29 hours during the interview week are assigned a code of "0 – less than 30 hours." Respondents who worked 30 or more hours that week are assigned a code of "1 – 30 hours or more."

DOI_RETIRED – DATE OF INTERVIEW STATUS - RETIRED
Current retirement status at the date of each interview was established by using available questions that allow respondents to specifically report retirement in a response category. There are only one or two such question opportunities in the NLSY79 in any given year, and they are not necessarily asked of the entire sample. For later survey years (1994-2016) comments entered during the interview are available. Comments for these later years were examined to identify respondents who reported current retired status outside the structure of the proscribed interview questions. In more recent survey years (using 2012 as an example) to create DOI_RETIRED, the following succession of questions is used:

2012 (round 25)
RETIRE_EXP_P2_1A = 5 (R is "retired")
 else QES-23A.## = 17 (R quit job due to being "retired")
 
In earlier survey years, (using 1990 and 1998 as examples) the following succession of questions is used to create DOI_RETIRED:

1990 (round 12)
QES-23A.## = 17 (R quit job due to being "retired")

1998 (round 18)
QES-23A.## = 17 (R quit job due to being "retired")
 else ESR_KEY = 5 (R is "not in labor force – retired")

DOI_DISABLED – DATE OF INTERVIEW STATUS - DISABLED
Similar to retirement status, current disability status at the date of each interview was established by using available questions that allow respondents to specifically report disability/inability to work in a response category. NLSY79 questionnaires offer various opportunities in any given year to report disability, although all are not necessarily asked of the entire sample. In more recent survey years (using 2012 as an example) to create DOI_DISABLED, the following succession of questions is used:

2012 (round 25)
RETIRE_EXP_P2_1A = 4 (R is "disabled and unable to work")
 else QES-23A.## = 10 (R quit job due to "health, disability or medical problems")
  else QES-43.##.## = 2 (R's reason for a current within-job gap is being "ill, disabled or unable to work")
   else Q7-19.## = 9 (R's reason for not looking for work during between-job gap is "ill-health, physical disability")
    else Q11-3 = 1 (R's health would keep R from working on a job for pay now)

In earlier survey years, (using 1990 as an example) the following succession of questions is used to create DOI_DISABLED:

1990 (round 12)
QES-23A.## = 10 (R quit job due to "health, disability or medical problems")
 else ESR-KEY = 6 (R's ESR status is "disabled")
  else QES-43.##.## = 2 (R's reason for a current within-job gap is being "ill, disabled or unable to work")
   else Q7-19.## = 9 (R's reason for not looking for work during between-job gap is "ill-health, physical disability")
    else Q11-3 = 1 (R's health would keep R from working on a job for pay now)
 
As mentioned for the retirement variables above, comments entered during the interview were available for survey years 1994-2016. Comments for these later years were examined to identify respondents who reported current disability status outside the structure of the proscribed interview questions as listed above.

Below is the *.sas code that created the 2016 Date of Interview Current Status variables:

COMPUTE 2016 (ROUND 27) DOI_EMPLOYED & DOI_HOURS_WORKED
 
array status   (i)      _1-_1983 STATUS_WK_NUM1984-STATUS_WK_NUM2082;
array hrs   (i)  _1-_1983 HRS_WORKED_WK_NUM1984-HRS_WORKED_WK_NUM2082;

do i=1984 to 2082;
    if i=CURINT_WK#_2016 then do;
        empstat_wbi=status;
    end;
end;

if empstat_wbi>0 & empstat_wbi<=5 then DOI_EMPLOYED=0;
else if empstat_wbi=7 then DOI_EMPLOYED=1;*in military service;
else if empstat_wbi>=100 then DOI_EMPLOYED=1;
else if empstat_wbi=0 then DOI_EMPLOYED=-3;*no info reported to account for week,;

if CURINT_WK#_2016=-4 then DOI_EMPLOYED=-5;

do i=1984 to 2082;
    if i=CURINT_WK#_2016 then do;
        hrs_doi=hrs;
    end;
end;

if DOI_EMPLOYED=0 then DOI_HOURS_WORKED=-4;
else if DOI_EMPLOYED=1 then do;
    if hrs_doi<0 & hrs_doi>-4 then DOI_HOURS_WORKED=-3;
    else if hrs_doi>=0 & hrs_doi<30 then DOI_HOURS_WORKED=0;
    else if hrs_doi>=30 then DOI_HOURS_WORKED=1;
end;
else if DOI_EMPLOYED=-3 then DOI_HOURS_WORKED=-3;
else if DOI_EMPLOYED=-5 then DOI_HOURS_WORKED=-5;

COMPUTE 2016 (ROUND 27) DOI_RETIRED & DOI_HOURS_DISABLED

*why leaving a job;
array stopd (*) EMPLOYER_STOPDATE.01~D EMPLOYER_STOPDATE.02~D EMPLOYER_STOPDATE.03~D EMPLOYER_STOPDATE.04~D EMPLOYER_STOPDATE.05~D;
array stopm (*) EMPLOYER_STOPDATE.01~M EMPLOYER_STOPDATE.02~M EMPLOYER_STOPDATE.03~M EMPLOYER_STOPDATE.04~M EMPLOYER_STOPDATE.05~M;
array stopy (*) EMPLOYER_STOPDATE.01~Y EMPLOYER_STOPDATE.02~Y EMPLOYER_STOPDATE.03~Y EMPLOYER_STOPDATE.04~Y EMPLOYER_STOPDATE.05~Y;
array qes23a (*)  QES-23A.01 QES-23A.02 QES-23A.03 QES-23A.04 QES-23A.05;
*within job gap;
array qes31d (*)  QES-31.01.01~D QES-31.01.02~D QES-31.01.03~D
             QES-31.02.01~D QES-31.02.02~D QES-31.02.03~D
             QES-31.03.01~D QES-31.03.02~D QES-31.03.03~D
             QES-31.04.01~D QES-31.04.02~D QES-31.04.03~D
             QES-31.05.01~D QES-31.05.02~D QES-31.05.03~D;
array qes31m (*) QES-31.01.01~M QES-31.01.02~M QES-31.01.03~M
             QES-31.02.01~M QES-31.02.02~M QES-31.02.03~M
             QES-31.03.01~M QES-31.03.02~M QES-31.03.03~M
             QES-31.04.01~M QES-31.04.02~M QES-31.04.03~M
             QES-31.05.01~M QES-31.05.02~M QES-31.05.03~M;
array qes31y (*) QES-31.01.01~Y QES-31.01.02~Y QES-31.01.03~Y
             QES-31.02.01~Y QES-31.02.02~Y QES-31.02.03~Y
             QES-31.03.01~Y QES-31.03.02~Y QES-31.03.03~Y
             QES-31.04.01~Y QES-31.04.02~Y QES-31.04.03~Y
             QES-31.05.01~Y QES-31.05.02~Y QES-31.05.03~Y;
array qes33 (*)  QES-33.01.01 QES-33.01.02 QES-33.01.03
             QES-33.02.01 QES-33.02.02 QES-33.02.03
             QES-33.03.01 QES-33.03.02 QES-33.03.03
             QES-33.04.01 QES-33.04.02 QES-33.04.03
             QES-33.05.01 QES-33.05.02 QES-33.05.03;
array qes43     (*)  QES-43.01.01 QES-43.01.02 QES-43.01.03
                     QES-43.02.01 QES-43.02.02 QES-43.03.03
                     QES-43.03.01 QES-43.03.02 QES-43.03.03
                     QES-43.04.01 dum dum
                     QES-43.05_01 dum dum;
*between job gaps;
array q711d (*) Q7-11.01~D Q7-11.02~D Q7-11.03~D;
array q711m (*) Q7-11.01~M Q7-11.02~M Q7-11.03~M;
array q711y (*) Q7-11.01~Y Q7-11.02~Y Q7-11.03~Y;
array q719  (*)  Q7-19.01 Q7-19.02 Q7-19.03;

***** retire or not at interview date;
if QES-23A.01=-5 then qes23aR=-5;
else do;
 qes23aR = 0;
 do i = 1 to dim(qes23a);
  if (qes23a(i) = 17) then qes23aR = (qes23aR + 1);
 end;
end;

if QES-23A.01=-5 then DOI_RETIRED=-5;
else do;
 DOI_RETIRED=0;
 if (RETIRE_EXP_P2_1A = 5) then DOI_RETIRED = 1;
 else if qes23aR > 0 then do;
  DOI_RETIRED = 1;
  flag_r1=1;
 end;
end;

***** disabled or not at interview date;
if QES-23A.01=-5 then do;
 qes23aD=-5;
 qes43D=-5;
 q719D=-5;
 q113D=-5;
end;
else do;
 qes23aD=0; qes23aD3=0;
 qes43D=0; qes43D3=0;
 q719D=0; q719D3=0;
 q113D=0;
 
 do i = 1 to dim(qes23a);
  if work=0 & qes23a(i) = 10 & ( (0<stopy(i) &  stopy(i)<curdate~Y) or (stopy(i)=curdate~Y & 0<stopm(i) & stopm(i)<curdate~M)
     or (stopy(i)=curdate~Y & stopm(i)=curdate~M & 0<stopd(i) & stopd(i)<curdate~D))  then qes23aD = qes23aD + 1;
 end;
 if qes23aD=0 then do i = 1 to dim(qes23a);
  if work=0 & qes23a(i) = 10 & (stopy(i) in (-1 -2 -3) or stopm(i) in (-1 -2 -3) or stopd(i) in (-1 -2 -3)) then qes23aD3= qes23aD3+1;
 end;
 
 do i = 1 to dim(qes43);
  if (qes43(i) = 2 or qes33(i)=9) & (qes31d(i)=curdate~D & qes31m(i)=curdate~M & qes31y(i)=curdate~Y) then qes43D = (qes43D + 1);
 end;
 if qes43D=0 then do i = 1 to dim(qes43);
  if (qes43(i) = 2 or qes33(i)=9) & (qes31d(i) in (-1 -2 -3) or qes31m(i) in (-1 -2 -3) or qes31y(i) in (-1 -2 -3)) & work=0 then qes43D3=qes43D3 + 1;
 end;
 
 do i = 1 to dim(q719);
  if q719(i) = 9 & (qes_b in (0 -4) or (q711d(i)=curdate~D & q711m(i)=curdate~M & q711y(i)=curdate~Y)) then q719D = (q719D + 1);
 end;
 if q719D=0 then do i = 1 to dim(q719);
  if q719(i) = 9 & (q711d(i)  in (-1 -2 -3) or q711m(i) in (-1 -2 -3) or q711y(i) in (-1 -2 -3)) then q719D3 = (q719D3 + 1);
 end;
 
 if q11_3=1 then q113D=1;
end;

if QES-23A.01=-5 then DOI_DISABLED=-5;
else do;
 DOI_DISABLED=0;
 if (RETIRE_EXP_P2_1A = 4) then DOI_DISABLED = 1;
 else if qes23aD>0 then do; DOI_DISABLED=1; flag_d1=1; end;
 else if qes43D>0 then do;  DOI_DISABLED=1; flag_d2=1; end;
 else if q719D>0 then do;   DOI_DISABLED=1; flag_d3=1; end;
 else if q113D>0 then do;   DOI_DISABLED=1; flag_d4=1; end;
 else if qes23aD3>0 then do; DOI_DISABLED=1; flag_d13=1; end;
 else if qes43D3>0 then do;  DOI_DISABLED=1; flag_d23=1; end;
 else if q719D3>0 then do;   DOI_DISABLED=1; flag_d33=1; end;
end;

***** HE based on check;
if pubid=8270 then do; DOI_DISABLED=0; flag_d23=.; end;

*** hand-edits from comments for respondents identifying as retired inserted here;

*** hand-edits from comments for respondents identifying as disabled inserted here;