Two sets of variables from the main NLSY79 data set are necessary to link the employers from one contiguous year to the next. Comparable variables exist for all years allowing a link to be established through all contiguous interview years during which the respondent reported working for a specific employer. This document uses 1989 and 1990 for the purpose of demonstration. Both backward and forward employer linking is possible. The procedure for linking one employer or all employers through years is similar, and expands with the number of employers and years involved. Users should note that with the advent of the CAPI-administered surveys, beginning in 1993, the procedure has been simplified slightly. Beginning in 1993, only one variable per employer instead of two exists which may reflect the employer number from the previous survey year, if applicable. In addition, the CAPI instrument ensures that the CPS employer, if any, will always be employer #1. Therefore, it is only necessary to check the appropriate variable in the first employer supplement to determine CPS status. Although the same basic procedure must be followed to link employers through years, the number of variables that must be accessed is reduced.
A first set of variables identifies the CPS employer from the most recent year (1990).
1) To identify CPS employer in 1990 (Work History acronyms in parentheses):
R 33400.00 (CPSJO901) Int Check
90 - Is Job # 1 Same As Current Job?
R 33540.00 (CPSJO902) Int Check 90 - Is Job # 2 Same As Current Job ?
R 33680.00 (CPSJO903) Int Check 90 - Is Job # 3 Same As Current Job ?
R 33820.00 (CPSJO904) Int Check 90 - Is Job # 4 Same As Current Job ?
R 33960.00 (CPSJO905) Int Check 90 - Is Job # 5 Same As Current Job ?
A "YES" answer (code 1) to one of these questions indicates that that employer was identified in the CPS section of the main questionnaire as the "CURRENT OR MOST RECENT" (CPS) employer and some basic information about that employer was collected in the CPS section for 1990. There should never be more than one CPS employer for any respondent in a given survey year. In most cases, the CPS employer will be JOB # 1, but not in all cases. Therefore, all employers (1-5) must be checked.
Once the 1990 CPS employer has been identified, the next step is to determine if the CPS employer is a continuation of an employer reported in the previous contiguous interview year (1989 in this case). A second set of variables is used to identify which employer is the same as the 1990 CPS employer (or other employer if applicable) in the previous interview year (1989 in this case).
These variables are:
2) To match or link any 1990 employer with its 1989 employer info (if any) (Work History acronyms in parentheses):
R 33307.00 (PREV901) Employer
Number From Info Sheet, Item 6 Matches, Job #1
R 33313.00 (PREV901) Employer Number From Info Sheet, Item 7 Matches, Job #1
R 33447.00 (PREV902) Employer Number From Info Sheet, Item 6 Matches, Job #2
R 33453.00 (PREV902) Employer Number From Info Sheet, Item 7 Matches, Job #2
R 33587.00 (PREV903) Employer Number From Info Sheet, Item 6 Matches, Job #3
R 33593.00 (PREV903) Employer Number From Info Sheet, Item 7 Matches, Job #3
R 33727.00 (PREV904) Employer Number From Info Sheet, Item 6 Matches, Job #4
R 33733.00 (PREV904) Employer Number From Info Sheet, Item 7 Matches, Job #4
R 33867.00 (PREV905) Employer Number From Info Sheet, Item 6 Matches, Job #5
R 33873.00 (PREV905) Employer Number From Info Sheet, Item 7 Matches, Job #5
This match or link is made by looking at two variables. (Each job has two comparable variables associated with it). These variables identify the number that a job was assigned in the previous interview year, if that job was reported. (INFO SHEET ITEM 6 lists the employers for whom the respondent was actually working at the date of the last interview. INFO SHEET ITEM 7 lists the employers the respondent mentioned during the previous interview but was not actually working for on the date of the previous interview.) If both of these variables contain a valid missing code (-4), then this job was not reported in the previous interview year (1989) and therefore can not be linked to any employer in the previous year. This is essentially a "new" employer, reported for the first time during the current survey year (or possibly an employer reported before the previous interview). If either of these variables (but not both) contains a valid number (1 or greater), this is the number of that job in the previous interview year. For example, if R(33400.) contains a "YES" (1) answer, then JOB #1 in 1990 has been identified as the CPS employer. One would then look at R(33307.) and R(33313.) to see if either of these variables contains a valid code. If R(33307.) contains a "2" for example, this would mean that the CPS employer for 1990 (JOB #1) is the same employer as JOB #2 from 1989. One could then attach information from JOB #1 in 1990 to information from the second employer (JOB #2) in 1989 as a continuing record of the respondent's experience with that employer. One could also check to see if this employer was the CPS employer in 1989 as well, by examining the 1989 variables which are comparable to the first set listed above. Using corresponding variables through contiguous interview years, information for an employer can be traced back through the first time the employer was ever reported by the respondent.
Forward linking can be accomplished in much the same way, using the same sets of corresponding variables listed above.
For example to link the 1989 CPS employer with its continuation (if any) in 1990, one would identify the 1989 CPS employer by looking at the set of 1989 variables comparable to the first set listed above. Once the 1989 CPS employer is identified, one would examine, the second set of variables for 1990 employers (the same as those listed above in this case), to determine if any of these jobs lists the number for the 1989 CPS employer as a "previous job". If one of these 1990 employer match variables contains the employer number of the 1989 CPS job, the match has been established. Information about the employer identified in 1990 could be attached to information about the 1989 CPS employer as a continuing record of the respondent's experience with that employer. For example, if employer number 1 (JOB #1) is identified as the 1989 CPS employer, one would need to search through the second set of variables listed above to see if any of them contained a "1" for previous employer. If one of the employer match variables for JOB #3 for instance, contains a "1", this would mean that JOB #3 in 1990 is the same as the CPS employer (JOB #1) in 1989.
This procedure works through contiguous survey years. For example, if a respondent was interviewed in all years from 1985-1990, a direct match can not be made between an employer reported in 1990 and one reported in 1985 without first establishing matches (or the lack thereof) through all intervening years. However if a respondent was interviewed in 1985, and not again until 1990, a link between employers reported in these two years would be accomplished in the same manner as that described above between 1989 and 1990, as there would be no intervening year(s) to interfere with a direct match.
Users should be aware of several characteristics of the data.
First, the NLSY79 employment history data are employer-based. All references to a "job" should be understood to be a reference to an employer. Information about work duties and positions and/or changes in duties or position performed or held during the respondent's tenure with a specific employer is collected as part of the record for that specific employer. For example, a respondent may regard him/herself as having held a number of "jobs" or positions with employer #1 (JOB #1). However, any information collected about these different positions would all be regarded as information about employer #1 (JOB #1).
In survey years 1979-1992, while the CPS employer is most often the first job listed in a given year, this is not always the case. Therefore, identification of the CPS employer (a common focus among researchers) requires that the "CPS employer identification" variables for each job be examined. Likewise, it is not always the case that a CPS employer in one year that has been reported before or since will be the CPS employer in other years. Likewise, a job which was not the CPS employer in the first year it was reported, may become the CPS employer in a later survey year. In survey years 1993-2004, the job order is determined by computer. The CPS employer for these survey years if one is reported, is always the first employer.
Beginning in 1994, there is no specific variable identifying the first job explicitly as the CPS employer. The only exceptions to the rule that JOB #1 is the CPS job are respondents currently serving in the active military services, who have no CPS employer. The following equations identify these active military respondents in 1994 through 1998 respectively, for whom employer #1 is not the CPS employer.
Dummy variables, comparable to the CPS job identification variables discussed above, can be created for 1994 through 2004 using these equations.
For survey years 1979-1992, the data file contains two possible variables to access in identifying previous employer number of a given job. In survey years 1993-2004, the previous employer rostering is computerized. For those years, the NLSY79 data file contains one variable identifying previous employer number.
Beginning with the 2000 NLSY79 data release, the Work History data are included with the main data in a single data file. Researchers interested in working with employment history data should see "Appendix 18: Work History Data" in this document for further information.