COMPUTER OPERATIONS REQUEST #553 To: WLS Staff From: Rob Warren Date: January 10, 1995 Updated: October 3, 1995 by Jennifer Sheridan Re: The 1992 WLS Employment History Section: Instrument and Codebook Design OVERVIEW The 1992 WLS instrument collected an employment history from all 1992 respondents. This history begins at the date of the 1975 interview and ends at the date of the 1992 interview. (For 1975 nonrespondents, the history begins in January 1975.) In 7923 cases, the 1992 respondent's employment history for all employment spells of at least six months during this period was observed; however, in 570 cases, where individuals experienced more than 4 employment spells lasting six months or longer during this period, some spells in the middle of the employment history were ignored. In other words, for all but 570 of the 1992 respondents, the 1992 WLS data file contains information about each of the 1992 respondents' employment spells which lasted six months or longer between 1975 and 1992. This memorandum is intended to explain the design of the instrument and to assist in the use of the documentation of the resulting data. Figure 1 (COR426--available in the paper copy of Appendix A) shows the several paths that 1992 respondents may have taken through the 1992 WLS employment history section. To understand the design of the employment history section a distinction between an "employment spell" and a "job" must be noted. An employment spell is an uninterrupted (except for seasonal workers) period of time working for the same employer, including self-employment. Jobs, on the other hand, are characteristic work responsibilities, often but not always associated with a formal job title, which may or may not change during the course of employment with a single employer; that is, during an employment spell. In the employment history section, we collected up to five employment spells, and if the spell lasted six months or longer, the first and last jobs the respondent held within that spell. FIRST EMPLOYMENT SPELL: 1975 RESPONDENTS For those who responded to the 1975 questionnaire, the 1992 instrument began by verifying the 1975 report of 1975 employment status. If the 1975 respondent said that our 1975 data were inaccurate, we accepted the 1992 report of the 1975 employment status for the purposes of conducting the 1992 interview and for constructing the final data file. For example, if the 1975 respondent said in the 1992 survey that he or she was employed in 1975, but our 1975 data indicated that he or she was not employed in 1975, then we accepted the 1992 report. If a 1975 respondent was employed at the time of the 1975 interview, then the first section of the employment history in 1992 asked about the employment spell which began before the month of the 1975 respondent's interview in 1975. If the 1975 respondent was not employed at the time of the 1975 interview, then the first section of the employment history in 1992 asked about the first employment spell which began after the 1975 interview and which lasted for six months or longer. In other words, 1975 respondents could potentially begin the employment history section of the 1992 interview in one of two ways: their 1975 employment spell (if employed at the time of the 1975 interview), or their first employment spell lasting six months or longer which began after the 1975 interview (if not employed at the time of the 1975 interview). FIRST EMPLOYMENT SPELL: 1975 NONRESPONDENTS 1975 nonrespondents began the employment section by reporting on their first full-time, civilian employment spell. If the end date of this first full-time spell was later than January 1975, then this employment spell became the initial spell in the history from 1975 to 1992. If the end date of this first full-time spell was prior to January 1975, then the 1975 nonrespondent was asked whether he or she was working in January of 1975. If so, then the employment spell at which the 1975 nonrespondent was working in January 1975 became the initial spell in the history. If the 1975 nonrespondent was not working in January 1975, then we used the first employment spell of six months or longer that began after January 1975. SECOND EMPLOYMENT SPELL Next, each 1992 respondent was asked about the employment spell lasting six months or longer following the one just discussed. Thus, the second employment spell asked about in the employment history section refers to either: 1) the first spell lasting six months or longer which began after the 1992 respondent left the employer he or she worked for in 1975 (for 1975 respondents employed at the time of the 1975 interview or for 1975 nonrespondents who were employed in January 1975); - or - 2) the second spell lasting six months or longer which began after the 1975 interview (for 1975 respondents not employed at the time of the 1975 interview or for 1975 nonrespondents who were not employed in January 1975). SUBSEQUENT EMPLOYMENT SPELLS Each 1992 respondent was then asked how many more employment spells lasting six months or longer they had experienced subsequent to the spells previously discussed. If they experienced only one such spell lasting six months or longer, then they were next asked about this spell. If they experienced more than one such spell lasting six months or longer, then they were asked about the next to last and last employment spells lasting six months or longer. Thus, for the 570 persons with more than four spells, the middle spells were ignored. Finally, if the 1992 respondent was in the midst of an employment spell lasting less than six months at time of the 1992 interview, this spell was asked about in the last portion of the employment history section. SUMMARY Each 1992 respondent was asked about his or her first two and last two employment spells lasting six months or longer (starting in 1975) and may have been asked about a spell which began less than sixth months prior to the 1992 interview. For all but 570 cases, this design captured 1992 respondents' entire employment histories between the dates of the 1975 and 1992 interviews. EMPLOYMENT PATTERN VARIABLE The variable "RF001JS" indicates the pattern of employment spells reported by 1992 respondents. The patterns and frequencies are depicted in Figure 2. The first column of this pattern variable corresponds to the 1992 respondent's employment spell at the time of the 1975 interview. A "1" in this column indicates that the 1992 respondent was employed in 1975 at the time of the 1975 interview and that he or she was asked about this employment spell in 1992. A "0" in this column indicates that the 1992 respondent was not employed at the time of the 1975 interview, or that the 1992 respondent was not interviewed in 1975. (Note that 1975 nonrespondents cannot have a "1" in the first column, since they were not interviewed in 1975.) For 1975 respondents, the second column of the pattern variable refers to the first employment spell which the 1992 respondent began after the 1975 interview. Among 1992 respondents who were employed at the time of the 1975 interview, this column refers to the first employment spell which began after the 1992 respondent ended their 1975 employment spell. Among 1992 respondents who were not employed at the time of the 1975 interview, this column refers to the first employment spell lasting six months or longer which the 1992 respondent began after the time of the 1975 interview. For 1975 nonrespondents, the second column of the pattern variable refers to the first employment spell, whether the 1975 nonrespondent was working in January 1975 or not; that is, the pattern variable does not distinguish the 1975 nonrespondents who were working in January 1975 from those who were not just by using the pattern variable. (The analyst must use the start date of the first employment spell of 1975 nonrespondents to determine who was working in January 1975.) A "0" in each of the first two columns of the pattern variable indicates that the 1992 respondent was not employed at the time of the 1975 interview (or was not interviewed in 1975) and had not experienced an employment spell lasting six months or longer since 1975. The third column of the pattern variable refers to the second employment spell which the 1992 respondent began after 1975. Because the instrument was designed to collect information about the first two and last two employment spells lasting six months or longer, no cases contain a "1" in both the first and third columns of the pattern variable. The fourth column of "RF001JS" refers to the next to last employment spell lasting six months or longer; the fifth column refers to the last employment spell lasting six months or longer. In eight cases, the 1992 respondent made an error when reporting the number of employment spells left after the first two, and as a result, the last employment spell is reported in the next-to-last spell section. For these eight cases, there is a "1" in the fourth column and a "0" in the fifth column. Finally, the sixth column of the pattern variable pertains to the 1992 respondent's current employment spell which had lasted for less than six months at the time of the 1992 interview. JOBS WITHIN EMPLOYMENT SPELLS Within each reported employment spell which lasted six months or longer, the 1992 respondent was asked to describe their job when they began that employment spell and--if the job changed--their job when they ended that employment spell. If the 1992 respondent reported a spell which had lasted less than six months at the time of the 1992 interview, then they were asked only about their job when they began this spell. Thus, the data file contains information about as many as five employment spells and as many as nine jobs. CURRENT/LAST EMPLOYMENT SPELL The 1992 respondent's "current or last" spell is defined as the employment spell which he/she last held or which he/she currently holds. The "current or last" spell may have been ascertained in any of the five employment spell sections. The only employment spells which could have lasted for less than six months are the employment spell reported in 1975 and the current employment spell. The "current or last" spell could appear in any of the four employment spell sections (J1 - J4), depending on the number of employment spells the 1992 respondent held prior to this current spell. The analyst must use the pattern variable "RF001JS" and/or the start and end dates of the employment spell to identify 1992 respondents whose tenure in a given employment spell is less than six months. The employment history sections, and the characteristics of 1992 respondents who appear in those sections, can be found in Figure 3. A description of all employment variables can be found in Figure 4. FIGURE 2: PATTERN VARIABLE FOR THE OCCURRENCE OF EMPLOYMENT SPELLS, 1975-1992 (RF001JS) Pattern Survey Code Section Frequency cdeghi cdeghi |||||| 1 000000 357 (NO WORK BETWEEN 1975-1992/93) 2 000001 8 3 010000 918 4 010001 17 5 011000 544 6 011001 14 7 011010 291 8 011011 22 9 011100 4 (R ERROR - SHOULD HAVE BEEN "7") 10 011101 3 (R ERROR - SHOULD HAVE BEEN "8") 11 011110 246 12 011111 21 13 100000 2689 14 100001 38 15 110000 1549 16 110001 39 17 110010 836 18 110011 42 19 110100 1 (R ERROR - SHOULD HAVE BEEN "17") 20 110110 807 21 110111 47 FIGURE 3 Pattern Variable Section Respondent Characteristics (RF001JS) J1-FIRST SPELL: 1975 Respondent employed in 1975 13-21 1975 Respondent not employed in 19753-12 1975 Nonrespondent 3-12 1975 Respondent or Nonrespondent 2 employed less than six months at first spell at time of 1992 interview 1975 Respondent or Nonrespondent 1 never employed J2-SECOND SPELL: 1975 Respondent employed in 1975 15-21 1975 Respondent not employed in 19755-12 1975 Nonrespondent 5-12 1975 Respondent or Nonrespondent 4, 14 employed less than six months at second spell at time of 1992 interview 1975 Respondent or Nonrespondent 1-3, 13 with less than 2 employment spells J3-THIRD SPELL: 1975 Respondent employed in 1975 17-21 1975 Respondent not employed in 19757-12 1975 Nonrespondent 7-12 1975 Respondent or Nonrespondent 6, 16 employed less than six months at third spell at time of 1992 interview 1975 Respondent or Nonrespondent 1-5, 13-15 with less than 3 employment spells J4-FOURTH SPELL: 1975 Respondent employed in 1975 20-21 1975 Respondent not employed in 197511-12 1975 Nonrespondent 11-12 1975 Respondent or Nonrespondent 8, 10, 18 employed less than six months at fourth spell at time of 1992 interview 1975 Respondent or Nonrespondent 1-7, 9, 13-17, 19 with less than 4 employment spells JC-CURRENT SPELL:J1 SPELL 2, 3, 13 J2 SPELL 4, 5, 14, 15 J3 SPELL 6, 7, 9, 16, 17, 19 J4 SPELL 8, 10, 11, 18, 20 1992 Respondent employed less than six12, 21 months at current spell at time of 1992 interview FIGURE 4 4/26/95 RF001JA - RF020JA 1975 job aspirations for 1992 Employment History (1975-1992/93) Summary Variables RF001JS Pattern of employment spells from 1975 to 1992/93. RF002JS Current (1992/93) employment status. RF003JS-RF004JS Flags for work between 1957-1992/93. RF005JS-RF009JS Number of employment spells, jobs held, total years worked and other summary variables. RF010JS-RF012JS Ever had a pension plan, health insurance, or left a job for health reasons. First employment spell (only/last): 1975 employer or first employer after 1975 if unemployed in 1975 RF001J1 - RF004J1 Start and end dates with first employer and full- part time. RF005J1 - RF021J1 Industry and occupation codes for first or only job with first employer. RF022J1 - RF030J1 Location, benefits, and job changes with first employer. RF031J1 - RF032J1 Start date for last job with first employer and full- part time. RF033J1 - RF048J1 Industry and occupation codes for last or only job with first employer. RF049J1 - RF053J1 Ending of employment with first employer. Second employment spell (second/last): First post-1975 employer (or second employer after 1975 if unemployed in 1975) RF001J2 - RF004J2 Start and end dates with second employer and full-part time. RF005J2 - RF021J2 Industry and occupation codes for first or only job with second employer. RF022J2 - RF030J2 Location, benefits, and job changes with second employer. (FIGURE 4 CONTINUED) RF031J2 - RF032J2 Start date for last job with second employer and full-part time. RF033J2 - RF048J2 Industry and occupation codes for last or only job with second employer. RF049J2 - RF053J2 Ending of employment with second employer. NOTE: If a respondent worked for more than 4 employers for six months or longer, the spells held between the second employer and the next to last (third) are skipped. EMPLOYMENT SPELLS WILL CONTAIN INFORMATION ABOUT EMPLOYMENT SPELLS OF 6 MONTHS OR LONGER UNLESS THE SPELL IS THE LAST IN THE SEQUENCE OF SPELLS, AND THE RESPONDENT HAD BEEN IN THIS LAST SPELL LESS THAN 6 MONTHS AT THE TIME OF THE INTERVIEW. Third employment spell (third/next to last/last): RF001J3 - RF004J3 Start and end dates with third employer and full- part time. RF005J3 - RF021J3 Industry and occupation codes for first or only job with third employer. RF022J3 - RF030J3 Location, benefits, and job changes with third employer. RF031J3 - RF032J3 Start date for last job with third employer and full- part time. RF033J3 - RF048J3 Industry and occupation codes for last or only job with third employer. RF049J3 - RF053J3 Ending of employment with third employer. Fourth employment spell (fourth/next to last/last): RF001J4 - RF004J4 Start and end dates with fourth employer and full- part time. RF005J4 - RF021J4 Industry and occupation codes for first or only job with fourth employer. RF022J4 - RF030J4 Location, benefits, and job changes with fourth employer. RF031J4 - RF032J4 Start date for last job with fourth employer and full-part time. RF033J4 - RF048J4 Industry and occupation codes for last or only job with fourth employer. RF049J4 - RF053J4 Ending of employment with fourth employer. (FIGURE 4 CONTINUED) Current/last employer: NOTE: For the majority of individuals the most recent employer has already been described in J1-J4. If a respondent worked for 4 or more employers for 6 months or longer and the last/current employment spell was held for less than 6 months, all four employment spells would contain data for four > 6 month employers and the current/last employer would contain the < 6 month employer information. IF PATTERN = 12 OR 21 USE ii VARIABLES ELSE J4, J3, J2, J1. RF001JC - RF004JC Start and end dates with current/last employer and full-part time. RF005JC - RF021JC Industry and occupation codes for first or only job with current/last employer. RF022JC - RF030JC Location, benefits, and job changes with current/last employer. RF031JC - RF032JC Start date for last job with current/last employer and full-part time. RF033JC - RF048JC Industry and occupation codes for last or only job with current/last employer. RF049JC - RF053JC Ending of employment with current/last employer. RF054JC Tenure with current/last employer. Current/Last Job Characteristics RG001JJ-RG005JJ How respondent obtained job. RG006JJ-RG022JJ Wages and rate of pay. RG023JJ-RG026JJ Company size and characteristics. RG027JJ-RG038JJ Job supervision. (Respondents received either Wright or Jencks items.) RG028JJ-RG032JJC. Jencks supervision items. RG033JJ-RG038JJE. Wright supervision items. RG039JJ-RG060JJ Job satisfaction and conditions. RG061JJ-RG065JJ Job importance.