1 data/final draft M… · Web viewEach Household Record Card (1) enumerated all persons currently living in the household; (2) recorded for each person: name, relationship to respondent,

NLS OF OLDER AND YOUNG MEN

USER’S GUIDE

A Guide to the 1966–1990 NLS of Older Men Dataand the 1966–1981 NLS of Young Men Data

Prepared for the U.S. Department of Labor by

Center for Human Resource ResearchThe Ohio State University

Columbus, Ohio2002

This publication is prepared in conjunction with contract #J-9-J-0-0090 with the Bureau of Labor Statistics, U.S. Department of Labor. The contents do not necessarily represent the official position or policy of the Department

of Labor.

2

AcknowledgmentsThe Older Men and Young Men cohorts of the National Longitudinal Surveys were initially sponsored by the Office of Manpower, Automation, and Training and later sponsored by the Bureau of Labor Statistics, U.S. Department of Labor. Surveys of these cohorts were conducted by the U.S. Census Bureau for the Center for Human Resource Research at The Ohio State University. Additional funding for the 1990 survey of Older Men was provided by the National Institute on Aging.

This document pays tribute to the initiators of the NLS, Herbert S. Parnes and Howard Rosen, who had the vision to see that the collection of longitudinal data on the labor force experience of American men and women could provide a basis for policies fostering more efficient and equitable labor markets. Appreciation is given to the current Principal Investigator for the NLS Original Cohorts, Randall Olsen. Special thanks go to the other members of the documentation and archival teams: Linda Boyer, Tiffany Chase, Eric Fischer, Pat Rhoton, Cheryl Segrist, and Jay Zagorsky.

Amanda McClainJennifer HeringEditors

Help and InformationThis manual is one of a set of user-oriented documents available to the NLS research community containing information on the Original Cohorts (Older Men, Young Men, Mature Women, and Young Women), the NLSY79, the Children of the NLSY79, and the NLSY97. Appendix A in this document contains a quick reference guide to essential information about the NLS and the Older and Young Men cohorts. Persons needing more information on the NLS in general or the NLS cohort groups are encouraged to obtain copies of the NLS Handbook. Summaries of research completed on the NLS over the past several decades are presented on-line in the NLS Annotated Bibliography of Research, located at www.nlsbibliography.org. Persons needing detailed information on the other cohorts should read the NLS of Mature Women User’s Guide, the NLS of Young Women User’s Guide, the NLSY79 User’s Guide, the NLSY97 User’s Guide, and the NLSY79 Child Handbook 1986–1990. Ordering information for NLS documentation and data sets is available from NLS User Services, 921 Chatham

iii

Lane, Suite 100, Columbus, OH 43221-2418; Voice (614) 442-7366, FAX (614) 442-7329, or e-mail at [email protected].

iv

Table of Contents

Chapter 1: Introduction......................................................................11.1 The NLS of Older and Young Men.........................................................31.2 The NLS................................................................................................31.3 Surveys of the Older and Young Men...................................................61.4 The Data on CD....................................................................................91.5 Organization of the Guide....................................................................9

Chapter 2: Sample Design and Attrition.............................................112.1 Sample Design...................................................................................132.2 Screening Process..............................................................................132.3 Sampling Process...............................................................................142.4 Interview Schedule & Fielding Periods...............................................152.5 Interview Methods & Target Universe................................................172.6 Reasons for Noninterview..................................................................202.7 Sample Representativeness & Attrition..............................................242.8 Sample Weights.................................................................................27

Chapter 3: Guide to the Men’s Data..................................................313.1 Survey Instruments and Other Documentation..................................333.2 Types of Variables..............................................................................373.3 NLS Codebook System.......................................................................413.4 CD-ROM Search Functions..................................................................49

Chapter 4: Topical Guide to the NLS Men...........................................514.1 Age.....................................................................................................534.2 Alcohol and Cigarette Use..................................................................584.3 Aptitude, Achievement & Intelligence Scores....................................594.4 Class of Worker...................................................................................614.5 Crime, Delinquency & School Discipline.............................................634.6 Discrimination....................................................................................644.7 Educational Status & Attainment.......................................................654.8 Family Background.............................................................................704.9 Fringe Benefits...................................................................................734.10 Geographic Residence & Environmental Characteristics..................754.11 Health..............................................................................................814.12 High School & College Surveys........................................................904.13 Household Composition....................................................................924.14 Income & Assets...............................................................................994.15 Industries.......................................................................................1054.16 Job Characteristics.........................................................................1114.17 Job Satisfaction & Work Attitudes...................................................1124.18 Labor Force Status.........................................................................1164.19 Marriage & Fertility........................................................................1234.20 Military...........................................................................................1284.21 Occupations & Occupational Prestige Indices................................1304.22 Pensions, Social Security & Retirement..........................................1334.23 Poverty Status & Public Assistance Support Sources.....................1404.24 Race, Ethnicity & Nationality..........................................................142

1

4.25 Training..........................................................................................1444.26 Wages............................................................................................1494.27 Work Experience............................................................................151

Appendices.....................................................................................163Appendix A: Quick Reference Guide.......................................................165Appendix B: Original Cohort Dictionary of Key Words............................171Appendix C: Unpacking Multiple Entries.................................................176

Index.......................................................................................179

List of Tables and Figures

Table 1.1.1 The NLS: Survey Groups, Sample Sizes, Interview Years & Status5Table 2.4.1 Sample Sizes, Retention Rates, and Fielding Periods.................16Table 2.5.1 Reasons for Exclusion from the Eligible Samples: The Original

Cohorts........................................................................................18Table 2.5.2 Types of Interviews by Residence Status: 1990 Older Men

Resurvey......................................................................................19Table 2.6.1 Constructed & Raw Coding Categories for the Reason for

Noninterview Variables................................................................21Table 2.6.2 Reasons for Noninterview: Older Men 1966–83.........................22Table 2.6.3 Reasons for Noninterview by Race: Older Men 1967–83...........22Table 2.6.4 Reasons for Noninterview: Young Men 1966–81........................23Table 2.6.5 Reasons for Noninterview by Race: Young Men 1967–81..........23Table 2.7.1 Cohort Characteristics by Race...................................................24Figure 2.7.1 Interview Completion Percentages by Race and Survey Year:

Older Men 1966–83......................................................................25Figure 2.7.2 Interview Completion Percentages by Race and Survey Year:

Young Men 1966–81....................................................................25Table 2.7.2 Number of Interviews Completed out of 13 Survey Years by Race:

Older Men 1966–90......................................................................26Table 2.7.3 Number of Interviews Completed out of 12 Survey Years by Race:

Young Men 1966–81....................................................................26Table 3.1.1 Question Numbering Conventions in PAPI Surveys.....................36Table 3.2.1 Reference Numbers by Survey Year...........................................38Figure 3.3.1 Sample Survey Question Codeblock..........................................42Figure 3.3.2 Sample Created Variable Codeblock.........................................43Table 4.1.1 Age, Date of Birth & Date of Death Variables: Older Men..........53Table 4.1.2 Ages of Interviewed Respondents by Survey Year (Unweighted):

Older Men....................................................................................54Table 4.1.3 Ages at Death & at Interview by Race (Unweighted): 1990 Older

Men..............................................................................................55Table 4.1.4 Age & Date of Birth Variables: Young Men.................................56Table 4.1.5 Ages of Interviewed Respondents by Survey Year (Unweighted):

Young Men...................................................................................56Table 4.2.1 Age at First and Last Use of Cigarettes (Unweighted): 1990 Older

Men..............................................................................................58

2

Table 4.3.1 Aptitude & Intelligence Tests: Young Men School Survey..........59Figure 4.4.1 Definitions of CPS Class of Worker Entries................................62Table 4.7.1 Young Men: Summary of Information on Educational Plans.......68Table 4.8.1 Birth Countries of Older Men Respondents’ Parents &

Grandparents...............................................................................70Table 4.8.2 Birth Countries of Young Men Respondents’ Parents &

Grandparents...............................................................................71Table 4.9.1 Numbers of Employed Respondents Reporting the Availability of

Various Types of Fringe Benefits at their Current Job: Young Men73Table 4.10.1 Created Variables for Geographic Residence and Mobility by

Survey Year: Older (O) and Young (Y) Men.................................75Figure 4.11.1. Older Men Health Question Series..........................................82Figure 4.11.2 Special Health Topics in 1990 Older Men Questionnaire.........83Table 4.11.1 Cause of Death of Older Men Respondents (1990)...................84Table 4.11.2 Reference Numbers for Older Men Psychological Well-Being

Questions.....................................................................................85Figure 4.11.3 Young Men Health Question Series..........................................88Table 4.13.1 Older Men Household Roster Questions 1966–90.....................93Table 4.13.2 Young Men Household Roster Questions 1966–81....................94Figure 4.13.1 Relationship Codes by Survey Year: Older and Young Men....95Table 4.13.3 Distribution of Respondents Living within Single & Multiple

Respondent Households: The Original Cohorts...........................96Table 4.13.4 Distribution of Respondents by Intra- & Inter-Cohort Households96Table 4.13.5 Number & Types of Dominant Pairs Identified during the Initial

Survey Years................................................................................97Table 4.14.1 Older Men Income Questions..................................................100Table 4.14.2 Young Men Income Questions.................................................100Table 4.14.3 Older and Young Men Asset Questions...................................101Table 4.14.4 Response Rates to 1981 Older Men Income and Asset Questions103Table 4.14.5 Response Rates to 1981 Young Men Income and Asset

Questions...................................................................................103Table 4.15.1 Industry of Current/Last Job by Type of Interview and Sample

Persons’ Labor Force Status: 1990 Older Men..........................106Table 4.15.2 Industrial Sector of Employed Respondents’ Current Job by

Number of Survey Years with an Industry: NLS of Older Men 1966–83.....................................................................................106

Table 4.15.3 Industrial Sector of Civilian Respondents’ Current/Last Job by Survey Year: Young Men 1966–81.............................................108

Table 4.15.4 Industrial Sector of Civilian Employed Respondents’ Current Job by Number of Survey Years with an Industry: Young Men 1966–81108

Table 4.15.5 Occupation & Industry Variables from the O & I Rewrite........109Table 4.17.1 Reference Numbers for Older Men Work Attitudes Questions 115Table 4.17.2 Reference Numbers for Young Men Work Attitudes Questions115Table 4.18.1 Employment Status Recode Creation: Older and Young Men 118Table 4.18.2 Employment Status Recode Codes: Older Men......................119Table 4.18.3 Employment Status Recode Codes: Young Men....................119Table 4.18.4 Labor Force Status: Older Men Civilian Respondents 1966–90120Table 4.18.5 Labor Force Status: Older Men Civilian Respondents by Race

1966–90.....................................................................................121Table 4.18.6 Labor Force Status: Young Men Civilian Respondents 1966–81122

3

Table 4.18.7 Labor Force Status: Young Men Civilian Respondents by Race 1966–81.....................................................................................122

Table 4.22.1 Older Men Pension Income Data.............................................133Table 4.22.2 Older Men Social Security/Railroad Retirement Income and

Disability Income Data...............................................................134Table 4.22.3 Reference Numbers for Older Men Retirement Attitudes

Questions...................................................................................137Table 4.23.1 Public Assistance Questions by Survey Year, Type of Assistance,

Reference Period & Universe: Older and Young Men................141Table 4.24.1 Race by Nationality: Older and Young Men............................142Table 4.24.2 Country Codes for the Nationality Variables...........................143Table 4.25.1 Numbers of Respondents Participating in Training Programs by

Survey Year and Race: Older and Young Men 1966–81............145Table 4.27.1 Number of Weeks Worked All Survey Years (Unweighted): Older

Men............................................................................................158Table 4.27.2 Number of Weeks Worked/Year by # of Survey Years Reported

Work (Unweighted): Older Men.................................................159Table 4.27.3 Number of Weeks Worked by Survey Year (Unweighted): Older

Men............................................................................................159Table 4.27.4 Number of Employers by Survey Year (Unweighted): Older Men159Table 4.27.5 Number of Weeks Worked All Survey Years (Unweighted): Young

Men............................................................................................160Table 4.27.6 Number of Weeks Worked Per Year by Number of Survey Years

Reported Work (Unweighted): Young Men................................160Table 4.27.7 Number of Weeks Worked by Survey Year (Unweighted): Young

Men............................................................................................160Table 4.27.8 Number of Employers by Survey Year: Young Men

(Unweighted).............................................................................161Table A.1 List of NLS Documentation..........................................................165Table A.2 Frequently Asked Questions........................................................170

4

Chapter 1: Introduction

1


1.1 The NLS of Older and Young MenComprising 5,020 men ages 45–59 in 1966, the NLS of Older Men cohort is the first of four original groups interviewed when the National Longitudinal Surveys (NLS) program began in the mid-1960s. The Older Men were followed by the Young Men, 5,225 respondents ages 14–24 in 1966. The U.S. Department of Labor selected these two cohorts to enable research on topics of particular interest in the labor market. Nearing the completions of their careers, the Older Men were making decisions about the timing and extent of their labor force withdrawal and planning for retirement. In contrast, the Young Men were completing school and entering the work force or joining the military, making initial career and job decisions that would impact their employment in the coming decades.

During the two-and-a-half decades following the original interview, the Older Men surveys focused on labor market decisions such as mid-career job changes, retirement expectations and experiences, timing of retirement, and reentry to the labor market after retirement. The surveys also regularly asked about respondents’ marital status, income and assets, and household composition. The early surveys collected background information on the respondents, such as educational attainment, training and military experiences, and characteristics of the respondents’ parents. In later years, as the respondents reached and passed retirement age, the interviews included more questions about topics such as health conditions, medical insurance, attitudes about retirement and life in general, and leisure activities.

Similarly, Young Men respondents answered questions about a number of topics. Various interviews have collected information not only about the respondents’ labor force behavior but also regarding education, training, military experiences, marriage, children, income and assets, health, and attitudes about work and family. A special survey of the high schools of Young Men respondents provides additional information about their educational experiences. The variety of questions and the long time span of the interviews allow researchers to track the experiences of two large group of men over large portions of their lives, permitting examination of many different research topics in a number of disciplines.

NLS of Older and Young Men User’s Guide 3


This detailed guide is designed for researchers working with the NLS of Older Men or the NLS of Young Men. Users who need general information about the NLS surveys should refer to the NLS Handbook. For those not familiar with the NLS, a brief overview of all the NLS cohorts follows.

1.2 The NLSThe National Longitudinal Surveys (NLS) are a set of six surveys sponsored by the Bureau of Labor Statistics (BLS), U.S. Department of Labor. These surveys have gathered information at multiple points in time on the labor market experiences of diverse groups of men and women. Each of the NLS samples consists of several thousand individuals, some of whom have been surveyed over several decades. The earliest NLS interviews began in 1966 under the original sponsorship of the Office of Manpower, Automation, and Training (now the Employment and Training Administration). These cohorts were chosen in an effort to understand specific issues pertaining to the U.S. labor market, such as retirement, the return of housewives to the labor force, and the school-to-work transition. Since that time, however, the content of the surveys has been expanded to provide useful information on an extremely broad range of topics.

The first four NLS cohorts (Older Men, Mature Women, Young Men, and Young Women) were selected in the mid-1960s because each faced important labor market decisions that were of special concern to policy makers. In 1977, a survey of all known NLS data users and the recommendations of a panel of experts convened by the Department of Labor led to two decisions. The first was to continue the surveys of the four Original Cohorts for an additional five years (as long as attrition did not become a problem). The second decision was to begin a new longitudinal study of a panel of young men and young women.

This new study was initiated to permit a replication of the analysis of the 1960s Young Men and Young Women cohorts and to assist in the evaluation of the expanded employment and training programs for youth legislated by the 1977 amendments to the Comprehensive Employment and Training Act (CETA). To these ends, in 1978 a national probability sample was drawn of young women and young men living in the U.S. and born between January 1, 1957, and December 31, 1964. This sample included an overrepresentation of blacks, Hispanics, and economically disadvantaged non-black/non-Hispanics. With

NLS of Older and Young Men User’s Guide4


funding from the Department of Defense and the Armed Services, an additional group of young persons serving in the military was selected for interviewing. This sample of civilian and military youth, called the National Longitudinal Survey of Youth 1979 (NLSY79), was first interviewed in early 1979 and has been re-interviewed regularly since that date.

The Children of the NLSY79 survey, begun in 1986, further enhanced the NLS. With funding from the National Institute of Child Health and Human Development (NICHD) and a number of additional government agencies and private foundations, detailed information on the development of children born to NLSY79 women has supplemented the data on mothers and children collected during the regular youth surveys. During these biennial surveys, a battery of child cognitive, socio-emotional, and physiological assessments are administered to NLSY79 mothers and their children. In addition to these assessments, the Children of the NLSY79 are also asked a number of questions in an interview setting. Beginning in 1994, children age 15 and older, the “Young Adults,” replied to a separate survey with questions similar to those asked of their mothers.

With the aging of the NLSY79 cohort, another longitudinal cohort has been started. The National Longitudinal Survey of Youth 1997 (NLSY97) collects information on the circumstances that influence or are influenced by the labor market behaviors of youth ages 12–16 as of December 31, 1996. Data on the youth’s educational experience, along with his or her family and community backgrounds, are also included in the survey. Documenting the transition from school to work, this survey is designed to be representative of the population born during the period 1980 to 1984.

Information on sample sizes, interview years, and the current survey status of each respondent group is presented in Table 1.1.1 below.



Table 1.1.1 The NLS: Survey Groups, Sample Sizes, Interview Years & Status

Survey Group Age Cohort Birth Year Cohort

Initial Sample Size

Initial Survey Year Survey Status

Older Men 45–59 (as of 4/1/66) 4/2/07–4/1/21 5020 1966 EndedMature Women 30–44 (as of 4/1/67) 4/2/23–4/1/37 5083 1967 ContinuingYoung Men 14–24 (as of 4/1/66) 4/2/42–4/1/52 5225 1966 EndedYoung Women 14–24 (as of 1/1/68) 1944–1954 5159 1968 ContinuingNLSY79 14–21 (as of 12/31/78) 1957–1964 12686 1979 Continuing NLSY79 Children Birth–14 — 3 1986 Continuing NLSY79 Young Adults 15 & older — 3 1994 ContinuingNLSY97 12–16 (as of 12/31/96) 1980–1984 9022 1997 Continuing

1 The 1990 interview surveyed both living respondents and next-of-kin of deceased respondents.2 NLSY79 Young Adult respondents were initially interviewed as part of the NLSY79 Children sample. Beginning in 1994, those 15

and older were surveyed separately. In 1998, youths older than age 20 were not interviewed.3 The sizes of the NLSY79 Children and Young Adult samples are dependent on the number of children born to NLSY79 respondents.

Since this number is still increasing, original sample sizes are omitted.

Administration of the Project. Responsibility for the administration of the NLS resides with the Bureau of Labor Statistics (BLS), an agency of the U.S. Department of Labor. Established in 1884, BLS is responsible for the analysis and publication of data series on employment and unemployment, prices and living conditions, compensation and working conditions, productivity, occupational safety and health, and economic growth and employment projections. Its mission is to promote the development of the U.S. labor force by gathering information about the labor force and disseminating it to policy makers and the public so participants in those markets can make more informed, and thus more efficient, choices.

The NLS program supports BLS in this mission. The surveys are part of a longitudinal research program that includes in-house analyses, extramural grants, and other special projects. The NLS program is housed in the Office of Employment Research and Program Development, headed by an Assistant Commissioner. Direction of the NLS program is the responsibility of the Director of National Longitudinal Surveys.

BLS contracts with the Center for Human Resource Research (CHRR) at The Ohio State University to manage the NLS Original Cohorts and NLSY79, to share in the design of the survey instruments, to disseminate the data, and to provide user services. The Census Bureau collects data for the Original Cohorts through an interagency agreement with BLS. Data collection for the NLSY79 and the NLSY79 Children samples is subcontracted to the National Opinion Research Center



(NORC) at the University of Chicago. BLS contracts with NORC to manage the NLSY97 cohort and to collect the data for this survey; CHRR is subcontracted for data dissemination, documentation, and variable creation.

The project is assisted in its efforts by the NLS Technical Review Committee. Meeting twice each year, committee members provide recommendations regarding questionnaire design, additional survey topics, potential research uses, methodological issues, data distribution, and user services. The committee is multidisciplinary, reflecting the wide range of social scientists utilizing NLS data.

Ultimate responsibility rests with BLS for overseeing all aspects of the work undertaken by the other organizations. For more information about the NLS program, contact:

The National Longitudinal SurveysBureau of Labor Statistics2 Massachusetts Avenue NE, Room 4945Washington, DC 20212-0001(202) 691-7410; FAX: (202) 691-6425e-mail: [email protected]: www.bls.gov/nls

This detailed guide is designed for researchers who are either working or planning to work with the National Longitudinal Survey of Older Men or the National Longitudinal Survey of Young Men. Users who are interested in a general overview should refer instead to the NLS Handbook.

1.3 Surveys of the Older and Young MenIn 1966, two groups of men were designated for interview as part of the NLS. The first group comprised 5,518 men ages 45–59 as of March 31, 1966; of these, 5,020 (91.0 percent) participated in the initial 1966 survey of the Older Men cohort. The second group, men ages 14–24 as of the same date, included 5,713 individuals; 5,225 (91.5 percent) of these were interviewed during the first Young Men survey in 1966. In both cohorts, blacks were oversampled to ensure that they would be represented in sufficient numbers for analyses.

Respondents in the older men’s cohort completed 12 interviews from 1966–83, with an additional interview sponsored by the National Institute on Aging occurring in 1990. Interviewers administered seven of these interviews



personally and conducted five by telephone; the third survey of this cohort, in 1968, was mailed to the respondents. The young men’s cohort participated in 12 surveys from 1966–81; interviewers administered eight of the surveys personally and conducted four by telephone. Paper-and-pencil instruments (PAPI) were used exclusively during all interviews of both cohorts. Interviews with both men’s cohorts have ceased. Further information is provided in the “Interview Schedule & Fielding Periods” section of this guide.

Due to the sample design, a number of respondents shared a household with another NLS Original Cohort respondent at the time of the first survey. These multiple respondent households, although not representative of all possible relationships, represent unique samples for a number of research topics. For more information about multiple respondent households, users should consult the “Sample Design” section of chapter 2 and the “Household Composition” section of chapter 4 in this guide.

The surveys have collected three basic types of information: (1) the respondent’s work and non-work experiences, training investments, schooling, family income and assets, physical well-being, and geographic residence; (2) background information on the respondent’s family and household composition, and his marital and fertility history; and (3) supplementary data specific to the age, sex, stage of life, and/or labor market attachment of this cohort (e.g., military experiences, retirement plans and experiences, volunteer work). Data elements for the Older and Young Men are briefly described in the following paragraphs:

Alcohol and cigarette use. For the Older Men, data were collected in the 1990 survey on lifetime consumption of alcohol and lifetime use of cigarettes.

Attitudes and aspirations. The surveys included questions on the respondent’s attitude toward his current job, women’s roles, and satisfaction with life. Interviews of the Older Men contained a variety of mental health and mental status scales. The Young Men provided information about their knowledge of the World of Work, occupational plans, educational aspirations, and fertility expectations.



Demographic and family background characteristics. Information was collected on each respondent’s race, date of birth, and nationality; his residence and family structure at age 14 or 15; and his parents’ birthplaces, parents’ life status, and father’s occupation and education.

Discrimination. During select survey years, respondents were asked varying questions on whether they had experienced sex, age, or race discrimination at work.

Educational status and attainment. Older Men provided only limited data about educational attainment, including highest grade completed and degrees and diplomas received. For Young Men, data are available on current school enrollment status, highest grade attended or completed, attainment of a high school diploma or GED, type of high school curriculum, field of study at college, amount of college tuition, types of financial aid, and highest degree received. In addition, a separate school survey collected information about the high schools attended by Young Men.

Geographic and environmental characteristics. Data about the respondent’s geographic residence include the region of residence (South/non-South) and whether current residence is in a Standard Metropolitan Statistical Area (SMSA). Mobility information is provided by a set of variables comparing the respondent’s current state or SMSA with his previous residence, current job, and birthplace. A third series of variables details characteristics of the respondent’s environment, including the size of the labor force and the unemployment rate for the labor market of the respondent’s current residence. Users should be aware that specific states and SMSAs are not available and that data for the 1990 Older Men are more limited

Health status and health insurance. Information was collected from both cohorts on the presence and duration of health conditions that limit the respondent’s labor market work, the presence of physical problems such as vision and hearing difficulties, overall state of health, and difficulty performing various activities. Older Men also answered questions about the need for assistance in personal care, transportation, bill paying, and getting about; use of the medical system; and health insurance.



Household composition. For each family member (early survey years) or household member (later survey years) living in the respondent’s household at the time of the survey, information was collected regarding relationship to the respondent, birth date or age, school enrollment status, highest grade completed, and work experience.

Income and assets. Income data include the amount of income received by the respondent, spouse, and other family members in the past calendar year or past 12 months. Asset information includes the total market value of property and/or business owned, savings and retirement accounts, stocks, bonds, and debts such as mortgages.

Labor market status and transitions. Data include current labor force status (activity during most of the survey week) as well as, for those employed, information on occupation, industry, class of worker, rate of pay, hours worked per week, and attitude toward current job.

Leisure activities and volunteer work. The Older Men surveys collected information on the types and amount of time spent on leisure activities, and on the extent of unpaid volunteer work.

Marital and fertility histories. Information is available on the respondent’s marital status at each survey date and changes in his marital status over time. Fertility information includes the respondent’s total number of children; the number of his children (under age 18) who lived in the household; and, for each child living in the household at the survey date, the child’s age, date of birth, and whether the child was the respondent’s son or daughter. The interviews also gathered information on the respondent’s dependents other than his wife.

Military experiences. Although respondents actually serving in the Armed Forces were not interviewed, information about past military experiences was gathered from respondents who had returned to the civilian population. Older Men answered limited questions regarding periods of service, training received, and income from Armed Forces pensions. The Young Men provided more detailed information regarding military branch and rank, training received, military occupation, and return to civilian life.



Retirement plans and experiences. Questions regarding retirement were asked only of the Older Men. Information collected includes retirement plans and experiences of respondents; presence of compulsory retirement plan at current job; expected age at retirement; eligibility for retirement benefits and number of years covered under Social Security/Railroad Retirement; characteristics of employer pension plans; retirement training programs; and attitudes and expectations about retirement.

Training. Questions have been asked regarding participation in, type of, and usefulness of occupational training programs including types of certificates and diplomas received.

1.4 The Data on CDNLS data are available to the public on CD-ROM at a nominal charge. Data for the NLS of Older Men (1966–90) and the NLS of Young Men (1966–81) are distributed on the Original Cohort Databases compact disc. This disc contains the longitudinal record of each respondent, as well as information from the 1968 school survey. The disc also includes data for the mature and young women. Researchers should contact NLS User Services for further information regarding this data release.

1.5 Organization of the GuideThe remainder of the guide is organized as follows:

Chapter 2 contains the technical information on the Older and Young Men samples. It includes information on sample sizes and retention rates, sampling design and fielding periods, sample representativeness and attrition, and weighting.

Chapter 3 provides the user with practical information on Older and Young Men data. This section describes how the data are collected and created, arranged on the CD-ROM, and presented in the documentation.

Chapter 4 presents summary discussions of sets of variables, arranged alphabetically by topic. Persons interested in reviewing, for example, variables that contain information on a respondent’s labor market status or geographic residence should turn to the respective topical section with that name. Each



topical section includes variable summaries, references to relevant survey instruments or documentation items, and cautionary notes to users about inconsistencies in the data.

Appendices are the fifth section of this guide. The appendices present a quick reference guide to the data and SAS and SPSS programs for separating responses to multiple-answer questions.


Chapter 2: Sample Designand Attrition

13

Chapter 2: Sample Design and Attrition

2.1 Sample DesignEach of the original NLS samples was designed to represent the civilian noninstitutionalized population of the United States at the time of the initial survey. The Older Men cohort includes individuals who were ages 45–59 as of April 1, 1966, and the Young Men cohort consists of respondents ages 14–24 as of the same date.

Each cohort is represented by a multi-stage probability sample originally drawn by the Bureau of the Census from 1,900 primary sampling units (PSUs) that had originally been selected from the nation’s counties and cities for the experimental Monthly Labor Survey, conducted between early 1964 and late 1966. A primary sampling unit consists of Standard Metropolitan Statistical Areas (SMSAs), counties (or parishes in some states), parts of counties (parishes), and independent cities. A total of 235 sample areas comprising 485 counties and independent cities were chosen to represent every state and the District of Columbia. From the sample areas, 235 strata were created of one or more PSUs that were relatively homogeneous according to socioeconomic characteristics. Within each of the strata, a single PSU was selected to represent the stratum. Finally, within each PSU, a probability sample of housing units was selected to represent the civilian noninstitutionalized population. Because the addresses for the sample frame came from the 1960 Census, respondents are covered by Title 13 confidentiality restrictions. Therefore, variables that link respondents to PSUs are not available to public users, making it impossible to identify respondents by city or state.

2.2 Screening ProcessThe initial sample of about 42,000 housing units for all four NLS Original Cohorts was selected and screening interviews took place in March and April 1966. Of this number, about 7,500 units were found to be either vacant, occupied by persons whose usual residence was elsewhere, changed from residential use, or demolished. On the other hand, about 900 additional units were found created within existing living space or changed from what had been nonresidential space. A total of 35,360 housing units were available for interview, from which usable information was collected for 34,662 households, for a completion rate of 98.0 percent.



The original plan called for using the initial screening to select all four Original Cohorts. However, after the sample members for the Older Men were chosen, the sample was rescreened in September 1966 before the initial interview of the Young Men. This decision was made because a seven-month delay between the screening and first interview seemed inordinate due to the mobility of Young Men in their late teens and early twenties. To increase efficiency, it was decided to stratify the sample for the rescreening by the presence or absence of a 14- to 24-year-old male in the household. The probability was high that a household that contained a 14- to 24-year-old in March would also have such a member in September. However, to insure that the sample also represented persons who had moved into sample households in the intervening period, a sample of addresses that previously had no 14- to 24-year-old males was also included in the screening operation. Since a telephone number had been recorded for most households at the time of the initial interview, every attempt was made to complete the short screening interview by telephone. The sample of households obtained through rescreening for young men was subsequently used to obtain the two samples of women ages 30–44, the Mature Women, and ages 14–24, the Young Women (Shea, Roderick, Zeller and Kohen 1971).

User Notes: During the screening process a large number of multiple respondent households were designated for interview; more than half of respondents in the Mature Women, Young Women, and Young Men cohorts and one third of respondents in the Older Men cohort originated from multiple respondent households (i.e., a household with at least one other respondent). For more information on multiple respondent households and on the types of relationships that existed between respondent pairs (e.g., spouse, sibling, etc.), see the “Household Composition” section of this guide.

2.3 Sampling ProcessThe sample was designed to provide approximately 5,000 respondents—about 1,500 blacks and 3,500 non-blacks—for each of the men’s cohorts. The men were sampled differentially within four strata: whites in predominantly white enumeration districts (EDs), blacks in predominantly black EDs, whites in predominantly black EDs, and blacks in predominantly white EDs. The sampling rate of households in predominantly black EDs was between three and four times



that for households in predominantly white EDs in order to meet the survey requirement of providing separate reliable statistics for black respondents; the sample design called for oversampling of blacks at twice the expected rate in the total population. An enumeration district is a geographical area considered to be an appropriate size for an interviewer to complete all the necessary interviews within a prescribed time frame.

Following the initial household interview and screening operation, 5,518 men ages 45–59 as of April 1, 1966, were designated to be interviewed. After rescreening, 5,713 young men ages 14–24 as of April 1, 1966 were designated for interview. Initial interviews with both of the men’s cohorts occurred in 1966. Among the individuals designated for interview, 5,020 or 91.0% of the Older Men and 5,225 or 91.5% of the Young Men were interviewed in 1966.

User Notes: Initially, 5,027 Older Men respondents were interviewed. When the data were reviewed, it was discovered that the data file contained 5,034 records and that 7 men had duplicate records. These men were dropped from the sample. Due to technical considerations related to the use of data tapes, survey staff did not remove their records from the data set but rather assigned a value of “not available” (–128 or –999) for all variables for all 14 records. Therefore, although the cohort contains 5,020 respondents, the data actually include 5,034 observations. In other words, although a given variable may include data for 5,034 observations, 14 of these are dropped cases and will have a value of “not available.” The dropped cases have the following identification numbers (R00001.): 693, 694, 809, 810, 903, 904, 1146, 1147, 1237, 1238, 3436, 3437, 5010, and 5011. For every variable except R00001., all of these cases have missing values.

2.4 Interview Schedule & Fielding PeriodsIn the initial survey plan, respondents from each of the four Original Cohorts were to be interviewed yearly over a five-year period. Due to cost considerations, it was decided after the second survey of the Older Men to survey the two older groups (Older Men and Mature Women) biennially rather than annually. Due to their greater mobility, the younger groups were to continue to be interviewed annually. Because of the usefulness of these data and the relatively small



sample attrition, a decision was made at the end of the first five-year period to continue the interviews for another five years. At this point, the interviewing pattern changed from a yearly personal interview to a 2-2-1 schedule; each respondent was contacted by phone approximately every two years, then again in person one year after the second phone interview. The 2-2-1 schedule was continued through 1981 for the Young Men and 1983 for the Older Men, when interviewing of both cohorts ceased. The Older Men or their proxies were then personally reinterviewed in 1990. Table 2.4.1 depicts the years in which each cohort was surveyed, the type of interview used, the fielding period, and the number and percent of respondents with completed interviews.

Table 2.4.1 Sample Sizes, Retention Rates, and Fielding Periods

Survey Year

Type of Interview Fielding Period Total

InterviewedRetention

RateaRetention Rate among Living Respondents

Older Men1966 Personal May – July 5020 100.0 100.01967 Personal May – July 4744 94.5 95.61968 Mail May – July 4648 92.6 95.11969 Personal July – September 4381 87.3 91.51971 Personal July – September 4175 83.2 90.31973 Telephone July – September 3951 78.7 88.71975 Telephone July – September 3732 74.3 87.21976 Personal July – September 3487 69.5 83.41978 Telephone July – September 3219 64.1 80.21980 Telephone July – September 3001 59.8 78.71981 Personal July – September 2832 56.4 76.31983 Telephone July – September 2633 52.5 74.81990 Personal October – December 2092b 41.5 89.9

Young Men1966 Personal October – December 5225 100.0 100.01967 Personal October – December 4790 91.7 91.91968 Personal October – December 4318 82.6 83.01969 Personal October – December 4033 77.2 77.71970 Personal October – December 3993 76.4 77.11971 Personal October – December 3987 76.3 77.21973 Telephone October – December 4014 76.8 78.01975 Telephone October – December 3977 76.1 77.41976 Personal October – December 3695 70.7 72.01978 Telephone October – December 3538 67.7 69.31980 Telephone October – December 3438 65.8 67.51981 Personal October – December 3398 65.0 66.8

a Retention rate is defined as the percent of base year respondents who were interviewed in any given survey year. Included in the calculations are deceased and institutionalized respondents as well as those serving in the military.

b Interviews were completed during 1990 with 2,092 surviving members of the original sample and with 1,341 widows and 865 next-of-kin of deceased respondents from whom data about the decedent was obtained. Table 2.5.2 below presents additional information on the 1990 universes.



User Notes: Although each of the personal interviews contains data of roughly the same degree of completeness, data gathered during the telephone interviews was not meant to update the longitudinal record of a respondent. Rather, the telephone interviews were intended to obtain a brief update of information on each respondent and to maintain sufficient contact so that the lengthier personal interview could be completed. The combination of fluctuating fielding periods and type of interview (i.e., personal or phone) may affect not only the probability of reinterview but also the reference periods of time-related questions.

There is another source of inconsistency with respect to time references. A given year’s survey instrument may use the previous calendar year as a reference period for some questions, while other questions will collect data for the year since last interview. Income data, for example, were often collected for the calendar year, corresponding to the time frame for a respondent’s tax records; employment data were usually collected for the year since the last interview.

2.5 Interview Methods & Target UniverseListings of respondents to be interviewed were generated by Census and distributed to its 12 regional offices. Current addresses and contact information were generated from information on the various Household Record Cards as well as through a postal check conducted by Census. Cases were assigned to interviewers who lived in the same geographic area as the respondent. For each respondent in their caseload, interviewers received copies of the questionnaire, respondents’ Household Record Cards, flashcards, and information booklets.

Interviewers were responsible for contacting each respondent in their caseload and for using additional local resources to locate respondents who had moved since the last interview. Respondents who had moved outside the geographic district of their original interviewer were assigned another interviewer unless there was no interviewer nearby. In the latter event, an effort was made to interview the respondent by telephone.

Each respondent to be interviewed was sent various materials designed to encourage continued participation. Advance letters thanking respondents for their



continued participation and informing them of the coming survey were mailed prior to each interview period. Fact sheets highlighting recent research findings from each cohort’s survey data were also provided. Respondents who initially refused to participate in a survey were sent refusal letters by the regional offices designed to encourage their continued participation and were once again contacted by local level interviewers to secure the interview.

While the type of survey, personal or telephone, determined the chief mode of contact, an alternate contact method was used for certain respondents. During a personal survey, for example, those respondents who lived long distances from the Census interviewer’s base of operation or those for whom the Census supervisor decided that another contact method was warranted were contacted by telephone. Survey instruments were written in English only; multilingual interpreters were made available by the regional offices to interviewers who needed them.

The average length of an interview varied depending on the type conducted, with personal interviews lasting from 50–60 minutes and telephone interviews averaging 20–25 minutes. No stipends were paid to respondents in the men’s cohorts for their participation in the NLS.

Respondents selected for interviewing each year were, with the exceptions noted below, those who had participated in the initial year interviews and who were alive, living within the United States at the interview date, and noninstitutionalized. Subsequent to the first year interview for each cohort, those respondents who had refused to be interviewed were dropped from the sample; respondents who had been noninterviews for reasons other than death or refusal for two consecutive years were also eliminated from attempted interviewing. This noninterview exclusion was not applied to those members of the Young Men cohort who were subsequently inducted into the Armed Forces. No interviews were attempted with this group while they were on active military duty. They were, however, retained in the sample and attempts were made to reinterview them as soon as they left active military service. Table 2.5.1 depicts reasons for exclusion from the eligible samples; tables in section 2.6 present reasons for noninterview across survey years.



Table 2.5.1 Reasons for Exclusion from the Eligible Samples: The Original Cohorts

Years Exclusion Reason in EffectOut-of-Scope Reason Older Men Young Men

InstitutionalizedAll years except the 1990 survey during which interviews were conducted with both institutionalized respondents and widows

All years

In the Armed Forces All years All yearsResiding outside the U.S. All years All years

Deceased

All years except the 1990 survey during which information on deceased respondents was collected from interviews with 1,206 widows (or other next-of-kin) of deceased respondents and with a select number of men who had been reported deceased at an earlier survey point

All years

Refusal during any one previous interview All years except 1990 All yearsDropped due to two consecutive noninterviews for reasons other than refusal, death, or membership in the Armed Forces

1968–831969–81 excluding Young Men enlisted in the Armed Forces

Congressional refusal1 1983 NA

1 Congressional refusal refers to a congressional representative requesting a respondent not be contacted again for an NLS survey after a respondent has completed one or more survey rounds.

Survey design and fielding procedures for the 1990 Older Men resurvey differed substantially from those employed during earlier interviews. Respondents in this cohort were last surveyed in 1983 and, if living, would have been between 69 and 83 years of age at the time of the 1990 interviews. It was expected that nearly half of the original cohort members could be deceased by the time of this resurvey.

The goal of the 1990 interviews was expanded to include obtaining information about the original cohort member regardless of his ability to respond. Questionnaires were designed for interviews with not only respondents, called “sample persons” for the purposes of this survey, but with widows of deceased sample members, or other next-of-kin in the absence of a widow. Information was to be collected not only on the labor market activities, retirement experience, and health of the respondent, whether living or deceased, but also on the widow’s work experience, household composition, and family income. Institutionalized sample persons or widows—those who resided in nursing homes, homes for the needy, mental institutions, correctional facilities and long stay hospitals—were to be interviewed. Staff members at institutions could be contacted to provide information on residency within and medical insurance coverage for long-term care facilities. To assess cognitive functioning, e.g., orientation to time and place, long-



term memory, and arithmetic ability, the Short Portable Mental Status Questionnaire (SPMSQ) was administered to all sample persons and widows.

Proxy interviews were encouraged in cases where the sample person was physically or mentally unable to participate. Eligible proxies, in order of preference, included a sample person’s wife, child, relative, friend, or neighbor, with those residing with or close to the respondent (e.g., in the same household, in the same community) preferred over those living some distance away. Finally, Census address information, which had last been updated in 1983, was supplemented by locator information from the Social Security Administration for both sample persons and their beneficiaries. These unusual procedures resulted in some information being collected from or about 4,298 (86%) of the original cohort members. Interviews were completed with 2,092 original sample persons (90% of those designated by Census as alive in 1990) and with 2,206 respondent widows or other next-of-kin (82% of those designated deceased before the interviews began). Table 2.5.2 summarizes response rates and types of interviews for the 1990 resurvey.

Table 2.5.2 Types of Interviews by Residence Status: 1990 Older Men Resurvey

Residence StatusTypes of Interview Total Noninstitutionalized Institutionalized Unknown

Sample Person (SP) or SP Proxy 2092 1954 60 78Sample Person 1899 1877 22 —Proxy for Sample Person 151 77 33 41Staff Member for Sample Person 5 — 5 —Sample Person & Proxy 37 — — 37

Widow or Widow Proxy 1341 1205 38 98Widow 1213 1201 12 —Proxy for Widow 112 4 24 84Staff Member for Widow 2 — 2 —Widow & Proxy 14 — — 14

Other Next-of-Kin 865 865 — —

Total Interviews 4298 4024 98 176

2.6 Reasons for NoninterviewA cumulative ‘Reason for Noninterview’ variable was created by CHRR for the full sample of respondents in each cohort. This created variable is a combination of: (1) the noninterview reasons provided by Census for the subset of respondents designated as eligible for interview in that survey year and (2) the reason for



noninterview assigned during a previous survey to out-of-scope respondents. Instructions to interviewers on how to code the reason a respondent was not interviewed appear within the cohort-specific Interviewer’s Reference Manuals.

The set of noninterview coding categories present during the initial survey years of each cohort was supplemented over the years with additional reasons for noninterview, and the meaning of existing categories was refined. For each cohort, Table 2.6.1 presents the raw coding categories and specifies the survey years during which each category was utilized.

In addition to the cumulative ‘Reason for Noninterview’ variable, the 1990 resurvey of Older Men data include two noninterview variables which reflect the reasons for noninterview for just those respondents for whom interviews were attempted that year. R06014. provides the reason for noninterview for sample persons with whom an interview was attempted, and R07114. provides the reason for noninterview of the Older Men respondent’s widow. The cumulative 1990 variable R06014.01 provides the reason for noninterview for all Older Men respondents and is comparable to the single variable included in the other interview years.

User Notes: The 'Reason for Noninterview' variables in the 1990 Older Men survey are constructed in a different manner than in other years. Because the survey included widows, as well as proxies for both living and deceased sample members and their widows, the coding categories were altered to reflect the different types of refusals by those other than the original sample person. Additionally, attempts were made to interview institutionalized respondents, a significant change from previous years. Finally, the 1990 reason for noninterview variables cannot be used to identify deceased sample members. Researchers who wish to identify all respondents known to be deceased as of the 1990 interview should use the variable “Identification/Residence Status of all 90 Interviews Survey Year: 1990” (R06015.90) Researchers may also use “Age of SP at Death Calculated from 90 Interview & Date of Birth” (R07075.). All respondents with a positive value for this second variable, as well as the 33 respondents with a value of –127 (don’t know), were deceased at the time of the 1990 interview.



The reason for noninterview coding categories depicted in the tables below were constructed from the raw coding categories as shown in Table 2.6.1. For example, the conceptual category “can’t locate” is the sum of codes “0” and “3” for the Older Men and codes “1” and “4” for the Young Men. Tables 2.6.2–2.6.5 depict the number of respondents in each cohort not interviewed by survey year, reason, and race. Additional detail on nonresponse rates and numbers by age in 1966 for the Older Men and Young Men cohorts is presented in Parsons (1987a, 1987b).

Table 2.6.1 Constructed & Raw Coding Categoriesfor the Reason for Noninterview Variables

Cohort, Code, & Survey YearsConstructed Older Men Young Men

Category Raw Coding Category a Code Years b Code YearsCAN’T LOCATE “Unable to locate [contact] R - reason not

specified”[0] All [1] All

“[Unable to locate R] - mover - no good address” [3] All [4] All

INTERVIEW IMPOSSIBLE

“[Unable to locate R] - mover - good address given but interview impossible to obtain (“moved to Germany” or “lives too far from PSU” - distance too great)” c

[1] All [2] All

“[Unable to locate R] - mover - good address given but unable to obtain interview after repeated attempts, etc.”

[2] All [3] All

“[Unable to locate R] - nonmover - unable to obtain interview after repeated attempts, etc.”

[4] All [5] All

“Temporarily absent” [5] All [6] All

“Other” [9] All [11] All

REFUSAL “Refusal” [7] All [9] All

“Congressional refusal” [13] 1983 — —

OUT OF SCOPE “In Armed Forces” [11] 1978–83 [7] All

“Institutionalized” [6] All [8] All

“Moved outside U.S. (other than Armed Forces)” [12] 1978–83 [13] 1978–81

DECEASED “Deceased” [8] All [10] All

DROPPED “Non-interview for two years, R dropped from sample”

[10] 1968–83 [12] 1969–81

a Specific instructions to Census interviewers on use of these coding categories can be found in the cohort-specific Interviewer’s Reference Manuals.

b Due to the change in fielding procedures for the 1990 resurvey of the Older Men, noninterview coding categories are markedly different for that year and are excluded from this chart. “All years” thus refers to the 1966–83 surveys. Persons interested in the 1990 noninterview variables should reference R06014.00 and R06014.01.

c Beginning in the 1978/1979 survey years when the separate “moved outside the U.S.” coding category was added as a reason for noninterview, this “unable to locate” coding category was to exclude those respondents who had moved outside the U.S.



Table 2.6.2 Reasons for Noninterview: Older Men 1966–83

Reason for NoninterviewSurvey

YearTotal

InterviewedTotal Not

InterviewedCan’t

LocateInterview

Impossible Refusal Out of Scope Deceased Dropped

1966 5020 0 0 0 0 0 0 a

1967 4744 276 44 65 107 0 60 a

1968 4648 372 55 25 159 1 132 a

1969 4381 639 42 77 249 13 234 241971 4175 845 33 48 305 17 399 431973 3951 1069 50 22 347 14 567 691975 3732 1288 26 25 390 13 741 931976 3487 1533 13 39 518 14 841 1081978 3219 1801 18 36 591 30 1004 1221980 3001 2019 10 10 626 32 1206 1351981 2832 2188 4 13 687 26 1307 1511983 2633 2387 4 13 687 26 1498 159

Note: Table based on R00635., R01120., R01156., R01639., R02547., R02693., R02862., R03739., R04076., R04488., and R05497.10. ‘Reason for Noninterview’ variables are available for 1990; because coding categories differ substantially from those in the 1967–83 surveys, noninterview data for 1990 are omitted here.

a Dropped from the sample because of two consecutive noninterviews was added as a coding category in 1969.

Table 2.6.3 Reasons for Noninterview by Race: Older Men 1967–83

Reason for NoninterviewTotal Not

Interviewed Can’t Locate Interview Impossible Refusal Out of Scope Deceased Dropped

Survey Year

Non-black Black Non-

black Black Non-black Black Non-


black Black Non-black Black

1967 198 78 23 21 42 23 94 13 0 0 39 21 a a

1968 262 110 25 30 16 9 135 24 0 1 86 46 a a

1969 459 180 25 17 52 25 207 42 7 6 154 80 14 10

1971 584 261 11 22 30 18 257 48 9 8 251 148 26 17

1973 732 337 22 28 14 8 292 55 9 5 356 211 39 30

1975 889 399 11 15 20 5 329 61 9 4 471 270 49 44

1976 1079 454 9 4 34 5 440 78 9 5 530 311 57 51

1978 1261 540 10 8 22 14 505 86 20 10 636 368 68 54

1980 1412 607 5 5 7 3 536 90 23 9 765 441 76 59

1981 1523 665 0 4 10 3 580 107 17 9 828 479 88 63

1983 1667 720 2 2 10 3 585 102 14 12 963 535 93 66

Note: Table based on R00023. (race), R00635., R01120., R01156., R01639., R02547., R02693., R02862., R03739., R04076., R04488., and R05497.10. ‘Reason for Noninterview’ variables are available for 1990; because coding categories differ substantially from those in the 1967–83 surveys, noninterview data for 1990 are omitted here.

a Dropped from the sample because of two consecutive noninterviews was added as a coding category in 1969.



Table 2.6.4 Reasons for Noninterview: Young Men 1966–81

Reason for NoninterviewSurvey

YearTotal

InterviewedTotal Not

InterviewedCan’t

LocateInterview

Impossible Refusal Out of Scopea Deceased Dropped

1966 5225 0 0 0 0 0 0 b

1967 4790 435 47 45 66 263 14 b

1968 4318 907 68 91 136 588 24 b

1969 4033 1192 113 84 190 732 35 381970 3993 1232 117 65 227 672 48 1031971 3987 1238 106 89 281 541 59 1621973 4014 1211 207 70 350 292 76 2161975 3977 1248 193 47 438 211 89 2701976 3695 1530 147 108 615 185 96 3791978 3538 1687 122 93 729 165 117 4611980 3438 1787 130 40 801 151 132 5331981 3398 1827 61 80 866 136 139 545

Note: Table based on R00644., R01169., R01751., R02342., R03059., R03925., R04130., R04389., R05480., R05967., and R06835.a Large numbers of Young Men remained out-of-scope for interviewing in 1967–71 due to enlistment in the Armed Forces. A

separate out-of-scope coding category, “moved outside the U.S.,” was added in 1978. Respondents who could not be interviewed in 1967–76 because their residence—either within or outside of the U.S—was too far away were coded within the “interview impossible” category. Out-of-scope counts for pre-1978 survey years thus may be understated.

b Respondents who had been noninterviews for two consecutive survey years due to reasons other than refusal or death were eliminated from the eligible sample beginning with the 1969 interview.

Table 2.6.5 Reasons for Noninterview by Race: Young Men 1967–81

Reason for NoninterviewTotal Not

Interviewed Can’t Locate Interview Impossible Refusal Out of Scopea Deceased Dropped

Survey Year

Non-black Black Non-



black Black Non-black Black

1967 310 125 21 26 30 15 50 16 200 63 9 5 b b

1968 632 275 30 38 56 35 108 28 424 164 14 10 b b

1969 811 381 55 58 50 34 148 42 515 217 22 13 21 171970 825 407 51 66 42 23 173 54 473 199 30 18 56 471971 789 449 52 54 51 38 212 69 359 182 36 23 79 831973 717 494 91 116 46 24 261 89 175 117 45 31 99 1171975 725 523 82 111 30 17 326 112 118 93 52 37 117 1531976 939 591 66 81 77 31 482 133 103 82 55 41 156 2231978 1012 675 53 69 36 57 574 155 95 70 65 52 189 2721980 1089 698 57 73 17 23 634 167 87 64 72 60 222 3111981 1165 662 31 30 46 34 694 172 78 58 77 62 239 306

Note: Table based on R00023. (race), R00644., R01169., R01751., R02342., R03059., R03925., R04130., R04389., R05480., R05967., and R06835.

a Large numbers of Young Men remained out-of-scope for interviewing in 1967–71 due to enlistment in the Armed Forces. A separate out-of-scope coding category, “moved outside the U.S.,” was added in 1978. Respondents who could not be interviewed in 1967–76 because their residence—either within or outside of the U.S—was too far away were coded within the “interview impossible” category. Out-of-scope counts for pre-1978 survey years thus may be understated.

b Respondents who had been noninterviews for two consecutive survey years due to reasons other than refusal or death were eliminated from the eligible sample beginning with the 1969 interview.



2.7 Sample Representativeness & AttritionThe retention rate for the Older Men as of the 1983 survey was 52.5 percent, or 2,633 of the original 5,020 respondents. In 1990, 41.5 percent or 2,092 of the original respondents completed an interview, along with 1,341 widows and 865 other next-of-kin; some information was obtained about 85.6 percent of the cohort members. The retention rate for the Young Men as of the final 1981 survey was 64.9 percent, or 3,398 of the original 5,225 respondents. Retention rate is defined as the percent of base year respondents who were interviewed in any given survey year; included in the calculations are deceased and other out-of-scope respondents (see Table 2.6.1 for definitions).

An analysis of selected characteristics of respondents interviewed in the tenth year samples of the Original Cohorts found that noninterviews had not seriously distorted the sample representativeness of any of the cohorts for the characteristics studied (Rhoton 1984). A second analysis of differential attrition among wealthy and non-wealthy subsamples of each of the four Original Cohorts found that non-wealthy respondents of each cohort showed a consistent tendency toward greater attrition (Rhoton and Nagi 1991). Among the three younger cohorts, almost all of the difference between wealthy and non-wealthy subsamples is accounted for by attrition reasons other than the death of the respondent. For year-by-year retention rates, consult Table 2.4.1 in the “Interview Schedule and Fielding Periods” section of this chapter.

In Table 2.7.1, the racial composition of interviewed respondents is presented for the base survey year (1966) and the most recent interview year for which data are available. This table also provides information on numbers of deceased respondents by race. Figures 2.7.1 and 2.7.2 characterize, by race, the percentages of the each of the original men’s samples who were interviewed at each survey point.



Table 2.7.1 Cohort Characteristics by Race

Race1 # of Int’d. Respondents Racial Composition Retention Number of Deaths2 Older Men

1966 1990 1966 1990 1990 as % of 1966 As of 1990Non-black 3600 1603 71.7 % 76.6 % 44.5 % 1807

Black 1420 489 28.3 % 23.4 % 34.4 % 886Young Men

1966 1981 1966 1981 1981 as % of 1966 As of 1981Non-black 3787 2622 72.5 % 77.2 % 69.2 % 77

Black 1438 776 27.5 % 22.8 % 54.0 % 62

1 See section in this guide on “Race, Ethnicity & Nationality” for details on race classifications. Respondent totals in this table are based on R00023. for the Older Men and R00023. for the Young Men.

2 Numbers are derived from R07075. for the Older Men and R06835. for the Young Men.

Figure 2.7.1 Interview Completion Percentages by Race and Survey Year: Older Men 1966–831

1 Due to differences in survey design, 1990 completion rates are not included here. See Tables 2.5.2 and 2.7.1 for more information.

Figure 2.7.2 Interview Completion Percentages by Race and Survey Year: Young Men 1966–81



Finally, Tables 2.7.2 and 2.7.3 present the number of interviews completed by respondents, broken down by race. In these tables, the “number who completed” column shows how many respondents completed exactly that number of surveys. The “cumulative %” column shows a cumulative total percent of those completing at least a given number of surveys rather than a percentage of those completing an exact number of surveys.



Table 2.7.2 Number of Interviews Completed out of 13 Survey Years by Race:Older Men 1966–90

All Respondents Non-Black Respondents Black RespondentsNumber of Surveys1

Number who completed Cumulative % Number who

completed Cumulative % Number who completed Cumulative %

13 1601 31.9 1223 34.0 378 26.612 944 50.7 657 52.2 287 46.811 316 57.0 223 58.4 93 53.410 174 60.5 115 61.6 59 57.59 240 65.2 163 66.1 77 63.08 280 70.8 200 71.7 80 68.67 201 74.8 151 75.9 50 72.16 233 79.5 164 80.4 69 77.05 230 84.0 151 84.6 79 82.54 267 89.4 169 89.3 98 89.43 216 93.7 155 93.6 61 93.72 169 97.0 118 96.9 51 97.31 149 100.0 111 100.0 38 100.0

Total 5020 3600 1420

Note: This table is based on R00635.01, R01120.01, R01156.01, R01629.01, R02541.01, R02689.01, R02857.01, R03739.10, R04076.10, R04488.10, R05496.10, and R06015.90. The number of respondents completing 13 interviews does not represent the total for whom data are available. The 1990 resurvey collected information about deceased respondents from widows or other next-of-kin.

1 Surveys completed in any year, not necessarily consecutive survey years.

Table 2.7.3 Number of Interviews Completed out of 12 Survey Years by Race:Young Men 1966–81

All Respondents Non-Black Respondents Black RespondentsNumber of Surveys1

Number who completed Cumulative % Number who

completed Cumulative % Number who completed Cumulative %

12 2084 39.9 1692 44.7 392 27.311 611 51.6 440 56.3 171 39.210 486 60.9 355 65.7 131 48.39 354 67.7 249 72.2 105 55.68 364 74.6 268 79.3 96 62.27 254 79.5 167 83.7 87 68.36 224 83.8 127 87.1 97 75.05 194 87.5 96 89.6 98 81.84 143 90.2 75 91.6 68 86.63 177 93.6 98 94.2 79 92.12 190 97.2 126 97.5 64 96.51 144 100.0 94 100.0 50 100.0

Total 5225 3787 1438

Note: This table is based on R00645., R01169.10, R01751.10, R02343., R03059.10, R03925.10, R04130.10, R04390., R05468.10, R05967.10, and R06835.10.

1 Surveys completed in any year, not necessarily consecutive survey years.



2.8 Sample WeightsThis section is divided into a discussion of the procedures used to develop sample weights and a section on the practical application of these weights. Before using NLS data in analysis, researchers should consult the practical usage discussion to determine when weighting of data is appropriate. Sample-based weights in each of the NLS cohorts were designed to reflect the underlying population in the year in which each cohort was originally surveyed. Individual case weights were assigned to yearly interviews in such a way as to produce group estimates which are demographically representative of each cohort’s base year population when used in tabulations.

Base Year Sampling WeightsPopulation data derived from the NLS are based on multi-stage ratio estimates. The first step was to assign each sample case a basic weight consisting of the reciprocal of the final probability of selection. This probability reflects the differential sampling by race within each stratum of the four cohorts.

The base year weights for all those interviewed were adjusted to account for the overrepresentation of blacks in the sample as well as for persons who were not interviewed in the initial survey. This adjustment was made separately for each of eight groupings for the Older Men (based on the four Census regions [Northeast, North Central, South, West] by urban-rural residence) and 24 groupings for the Young Men (based on the four Census regions, race [non-black/black], and three place of residence groupings [urban, rural farm, and rural nonfarm]).

In the first stage of ratio weight adjustment, differences at the time of the 1960 Census between the distribution by race and residence of the population as estimated from the sample PSUs and that of total population in each of the four major regions of the country were taken into account. Using 1960 Census data, estimated population totals by race and residence for each region were computed by appropriately weighting the Census counts for PSUs in the sample. Ratios were then computed between these estimates (based on sample PSUs) and the actual population totals for the region as shown by the 1960 Census.

In the second stage ratio adjustment, sample proportions were adjusted to independent current estimates of the civilian noninstitutionalized population by



age, sex, and race. These estimates were prepared by carrying forward the most recent Census data (1960) to take account of subsequent aging of the population, mortality, and migration between the United States and other countries (Census Bureau 1966). The adjustment was made by race within three age groups for the Older Men and in four age groups for the Young Men.

Sampling Weight Nonresponse AdjustmentSubsequent to the initial interview of each cohort, reductions in sample size occurred because of noninterviews. To compensate for these losses, the sampling weights of the interviewed individuals were revised. Each cohort of the NLS consists of a panel of individuals in which no new individuals were permitted to enter after the base year. As a result, all reweighting of the sample after the initial survey round was calibrated to base year population parameters. This revision was done in two stages. First, out-of-scope noninterviews in each of the years were identified by Census and eliminated from the sample of noninterviews. This group consisted of individuals who were institutionalized, who had died, who were members of the armed services, or who had moved outside the United States, i.e., individuals who were no longer members of the noninstitutionalized civilian population of the United States.

The second stage in the adjustment acknowledged the possible nonrepresentative characteristics of the in-scope interviews. For each survey year, those who were eligible but not interviewed, as well as those who were interviewed, were distributed into nonresponse adjustment cells. For the Older Men, there were 24 nonresponse adjustment cells based on 1966 data regarding race (black and non-black), length of time in residence at first interview (nine or fewer years, ten or more years, N/A), and education (N/A, eight or fewer years, nine to eleven years, twelve or more years). The Young Men cohort was divided into 30 nonresponse cells based on 1966 data using the same race and residence variables as above, but with father’s occupation (white collar, service, blue collar, farm, N/A) instead of the education variables used with the Older Men. Within each of the cells, the base year sampling weights of those interviewed were increased by a factor equal to the reciprocal of the reinterview rate (using base year weights) in that year.



For the Young Men cohort, the sampling weights of those interviewed were further adjusted to account for the return to the civilian population of men who were in the armed services at the time of initial interview. This final adjustment made use of the first stage estimates described above and independent Census Bureau estimates of the civilian population by selected age categories and race.

Practical UsageThe men’s cohorts are based upon stratified, multi-stage random samples with oversamples of blacks. Each case in each interview year is assigned a weight specific to that year. This weight can be interpreted as an estimate of the number of people in the population of interest that the individual in the sample represents. The following is a discussion of the ramifications of the weights when used for data analysis.

To tabulate characteristics of the sample (sample means, totals, or proportions) for a single interview year in order to describe the population being represented, it is necessary to weight the observations using the weights provided. For example, to estimate the average hours worked in 1976 by men born in 1957 through 1964, researchers would simply use the weighted average of hours worked, where weight is the 1976 sample weight. These weights are approximately correct when used in this way, with item nonresponse possibly generating small errors. Other applications for which users may wish to apply weighting, but for which the application of weights may not correspond to the intended result, include:

Samples Generated by Dropping Observations with Item Nonresponse: Often users confine their analysis to subsamples for which respondents provided valid answers to certain questions. In this case, a weighted mean will not represent the entire population, but rather those persons in the population who would have given a valid response to the specified questions. Item nonresponse because of refusals, don’t knows or invalid skips is usually quite small, so the degree to which the weights are incorrect is probably quite small. In the event that item nonresponse constitutes a small proportion of the variables under analysis, population estimates (i.e., weighted sample means, medians and proportions) would be reasonably accurate. However, population estimates based on data items that have relatively high nonresponse rates, such as family



income, may not necessarily be representative of the underlying population of the cohort under analysis.

Data from Multiple Waves: Because the weights are specific to a single wave of the study, and because respondents occasionally miss an interview but are contacted in a subsequent wave, a problem similar to item nonresponse arises when the data are used longitudinally. In addition, occasionally the weights for a respondent in different years may be quite dissimilar, leaving the user uncertain as to which weight is appropriate. In principle, if a user wished to apply weights to multiple wave data, weights would have to be recomputed based upon the persons for whom complete data are available. If the sample is limited to respondents interviewed in a terminal or end point year, the weights for that year can be used.

Regression Analysis: A common question is whether one should use the provided weights to perform weighted least squares when doing regression analysis. Such a course of action may lead to incorrect estimates. If particular groups follow significantly different regression specifications, the preferred method of analysis is to estimate a separate regression for each group or to use dummy (or indicator) variables to specify group membership. If one wishes to compute the population average effect of, for example, education upon earnings, one may simply compute the weighted average of the regression coefficients obtained for each group, using the sum of the weights for the persons in each group as the weights to be applied to the coefficients. While least squares is an estimator that is linear in the dependent variable, it is nonlinear in explanatory variables, and so weighting the observations will generate different results than taking the weighted average of the regression coefficients for the groups. The process of stratifying the sample into groups thought to have different regression coefficients and then testing for equality of coefficients across groups using an F-test is described in most statistics texts.

If one is unsure of the appropriate grouping, one should consult a statistician or other person knowledgeable about the data set before specifying the regression model. Note that if subgroups have different regression coefficients, a regression on a random sample of the population would be misspecified.



ReferencesCensus Bureau. Current Population Reports. Series P-25, No. 352, 1966.

Parsons, Donald O. “Summary Tables: The National Longitudinal Surveys of Older Males 1966–1983.” Columbus, OH: CHRR, The Ohio State University, 1987a.

Parsons, Donald O. “Summary Tables: The National Longitudinal Surveys of Young Males 1966–1981.” Columbus, OH: CHRR, The Ohio State University, 1987b.

Rhoton, Patricia. “Attrition and the National Longitudinal Surveys of Labor Market Experience: Avoidance, Control, and Correction.” Columbus, OH: CHRR, The Ohio State University, 1984.

Rhoton, Patricia and Nagi, Karima. "Attrition by Wealth in the Original NLS Cohorts." Columbus, OH: CHRR, The Ohio State University, 1991.

Shea, John R.; Roderick, Roger D.; Zeller, Frederick A.; and Kohen, Andrew I. Years for Decision, Volume 1: A Longitudinal Study of the Educational and Labor Market Experience of Young Women. Manpower Research Monograph 24, vol. 1. Washington, DC: U.S. Government Printing Office, 1971.


Chapter 3: Guide to theMen’s Data

37

Chapter 3: Guide to the Men’s Data

This chapter provides some practical information about how NLS variables are collected, created, and arranged on the CD-ROM along with accompanying hard copy and electronic documentation. The first section describes the different survey instruments used to collect the raw Older and Young Men data during the field period. This section also explains how question numbers were assigned during the various survey years. Next, the guide discusses the primary types of NLS variables and the process by which each was assigned a reference number and title that serve to identify it throughout the NLS documentation system. The third section reviews the codebook—that is, the information about each variable contained on the CD-ROM—and the accompanying paper documentation. This discussion will help users understand how to interpret the various pieces of data presented in the NLS documentation system. Finally, this chapter gives researchers some basic instruction on using the search functions on the CD-ROM to find variables relating to the area of interest.

3.1 Survey Instruments and Other DocumentationThe primary variables found within the main data set of each NLS cohort were derived directly from one or more survey instruments (e.g., questionnaires, household interview forms, etc.). This section provides information on the conventions utilized within the NLS documentation system to identify questionnaire items from some of the primary survey instruments.

Certain other documents, namely Interviewer Reference Manuals and flowcharts, provide background information on how specific survey instruments were administered or offer the researcher additional tools for working with a specific survey instrument. While not actually survey instruments, these additional documents are described within this section.

Survey InstrumentsA unique set of survey instruments was used during each survey year to collect information from respondents. The term “survey instrument” refers to: (1) the questionnaires, which serve as the primary source of data on a given respondent, and (2) documents such as the household interview forms or household record cards that collect information on members of the respondent’s household.

Note that, while the source of the majority of variables in the main NLS data sets is the questionnaire or one of the other survey instruments, certain NLS variables



are created either from other NLS variables or from information found in an external data source.

Household Record CardsNLS surveys included the collection, during each interview, of information on the members of each respondent’s household. These data were collected primarily through the “Household Roster” section of the questionnaire, which in turn relied upon information provided by Census personnel on the separate Household Record Cards. Respondents were selected on the basis of a screening of sample households. Both the instruments used for the household data collection and the household screening instruments that were used to draw the samples of respondents are described below.

Household Screener & Household Record Cards: Prior to most surveys, Census interviewers completed or updated information found on a Household Record Card. Part of this information was transferred during the main interview to the “Household Roster” section of the questionnaire. The first Household Record Card (LGT-1, dated 2/23/1966) is the screening instrument used to select the Older Men and Young Men respondents for interview. Information for this first card was gathered from any available household member, while respondents provided comparable information in subsequent surveys. Each Household Record Card (1) enumerated all persons currently living in the household; (2) recorded for each person: name, relationship to respondent, whether this person is considered a household member (CPS definition), marital status, birth date, and sex; (3) summarized changes since the last survey in household composition; and (4) provided information on the respondent’s current and/or permanent address and telephone number at the time of interview, as well as the names of people who would know how to contact the respondent at the time of the next interview.

Several versions of the paper Household Record Cards were used, each covering approximately three surveys. The 1990 resurvey of the NLS of Older Men used a unique card, RMM-1, which reflected both sample person and widow information. While information from these cards does not, in general, appear as variables within any of the Original Cohort data files, certain information present on the cards detailing each respondent’s current household composition was transferred to the “Household Roster” section of the questionnaires. In addition, certain



respondent-specific demographic variables as of the initial survey year, notably age, birth date, race, and sex, were derived from the 1966 household screenings. Users should consult each survey’s Interviewer’s Reference Manual for the specific instructions and definitions used to complete each card.

QuestionnairesThere were separate and distinctly different questionnaires for each survey year. Each questionnaire was organized around a set of topical subjects, the titles of which usually appear on either the first page of each section of the questionnaire or as page headers.

User Notes: The questionnaires are critical elements of the NLS documentation system and should be utilized by each researcher to ascertain the wording of questions, coding categories, and the universe of respondents asked to respond to a given question.

Each questionnaire collected two general types of information: (1) information on the actual interview (e.g., interview dates, times, and contact methods) and (2) information supplied by the respondents on various topics related to their work and life experiences. Each survey instrument was organized around core sets of questions: current labor force status, retrospective work history, attitudes, health, marital history, household composition, assets, and income. In addition, the interview schedules contain special sets of questions on a variety of topics specific to the particular stage of life: retirement and pension plans for the Older Men and education, high school/college experiences, and future job plans for the Young Men.

Information SheetInformation Sheets (or flap items), located within the questionnaires, were usually designed in such a way that the interviewers could fold the sheet out to the side of the actual questionnaire and refer to the items on the flap during the interview. Various information items from previous interviews were clerically entered by Census and used by the interviewer during the interview. These included information such as name of previous employer, actual date of previous interview, marital status at the time of previous interview, and place of residence at the time of previous interview.



The interviewer also transcribed information recorded in the questionnaire during the current survey onto the Information Sheet. The only current survey year item that a user would need from the flap was “current marital status,” transcribed from the Household Record Card in certain survey years. Items frequently referenced during the interview were more conveniently located when transferred to the flap.

Questionnaire Item or Question Number: The questionnaire item or question number is the generic term referring to the printed source of data for a given variable. A questionnaire item may be a question, a check item, or an interviewer’s reference item that appears within one of the survey instruments. Each questionnaire item has been assigned a number or a combination of numbers and letters within the documentation system to help the user link each variable to its location in a survey instrument. In general, questionnaire items found in the NLS documentation reflect the printed question numbers in the questionnaire.

Four different designations were used within the documentation system to identify varying types of questionnaire items. These designations, depicted in Table 3.1.1, utilize the preprinted numbers and letters that appear within each survey year’s questionnaire; unnumbered questionnaire items reference the page number. Users can locate most variables in the questionnaire by finding the question number that appears to the right of each variable description within the codebook. CD-ROM users can access variable titles and codebook information via the “Accessing Data by Question Number” function.

Table 3.1.1 Question Numbering Conventions in PAPI Surveys

Question: Question Number 112E; 59E

Interview Check: Check Item (CH) CHJ3; CH AA

Interviewer Reference Item: Interviewer Reference (R) 123R; R4

Unnumbered Questions: Page Number PG1

In the vast majority of cases, the reference is to a specific question item found in the survey (e.g., 22F or 3B). The convention “CH” is used to identify interviewer check items that occur within the survey (e.g., CH B). Their purpose is to direct



the interviewer to the next appropriate question. The convention “R” denotes a reference item (e.g., R2 or 12R). Typically, reference items are grouped in a section of the survey instrument called the Information Sheet, which contains information that interviewers frequently refer to during the course of an interview. Items designated “R” in the survey instruments are also designated “R” in the documentation. Finally, when an item does not include a question number, only the page number (“PG”) of the questionnaire on which a particular item appears is identified (e.g., PG 1). The first page of most questionnaires contains unnumbered interview status information and transcribed Household Record Card information.

The absence of a question number entry or page number in the documentation indicates that a variable was not directly collected in the questionnaire and is therefore a created variable. In the sample codebook entry in Figure 3.3.2, later in this chapter, there is no entry in the questionnaire item space, indicating that this variable is not picked up directly from the interview schedule.

School SurveyA supplemental survey of the last secondary school attended by respondents within the Young Men and Young Women cohorts was conducted in 1968. This special survey was mailed to the designated high schools and was designed to collect academic performance information and intelligence scores for respondents, as well as information on the programs and facilities of each high school. The instrument was called the “Survey of Work Experience of Young Men and Women School Survey.”

Interviewer’s Reference Manual/Field Representative’s ManualEach survey instrument that goes into the field is accompanied by an Interviewer’s Reference Manual (or Field Representative’s Manual), which provides Census interviewers with background information on the NLS, respondent locating instructions, and detailed question-by-question instructions for coding/completing the questionnaire and Household Record Cards. Note that Interviewer’s Reference Manuals do not always include all the actual questions.

FlowchartsThe questionnaires are lengthy and often present the researcher with the complex task of determining the universe of respondents asked a specific



question. To assist in this task, flowcharts have been developed that graphically depict the skip patterns for some questionnaires. Hard copy flowcharts are available for some post-1977 surveys; comparable information for earlier questionnaires appears within the codebook under the heading “Universe Information.”

3.2 Types of VariablesFour types of variables are present in Older and Young Men data files. The type of variable affects: (1) the title or variable description which names each variable and (2) the physical placement of the variable within the codebook. Types of variables include:

1. Direct raw responses from a questionnaire or other survey instrument.

2. Edited variables constructed from raw data according to consistent and detailed sets of procedures (e.g., occupational codings, *KEY* variables, etc.).

3. Constructed variables based on responses to more than one data item either cross-sectionally or longitudinally and edited for consistency where necessary.

4. Variables provided by the Census Bureau or another outside organization based on sources not directly available to the user (e.g., characteristics of respondents’ geographical areas).

User Notes: In general, CHRR does not impute missing values or perform internal consistency checks across waves. Data quality checks most often occur in the process of constructing cumulative and current status variables such as ‘Highest Grade Completed.’

Reference NumbersEvery variable within the main NLS data set has been assigned an identifying number that determines its relative position within the data file and documentation system. Persons contacting NLS User Services should be prepared to discuss their question or problem in relationship to the reference number(s) of the variable(s) in question.

Reference numbers, once assigned, remain constant through subsequent revisions of the files. Reference numbers are assigned sequentially, with variables from the first survey year having a lower reference number than those



variables specific to the second year, and so forth. Occasionally, variables are created sometime after the year in which the data were actually collected. These variables are frequently given a reference number that reflects the year in which the actual data were gathered rather than the year the created variable was constructed.

Table 3.2.1 lists reference numbers for each survey year since 1966 for the Older and Young Men.

Table 3.2.1 Reference Numbers by Survey Year

Survey Year Older Men Reference Numbers Young Men Reference Numbers1966 R00001.–R00585. R00001.–R00633.

1967 R00635.–R01075. R00635.–R01149.

1968 R01100.–R01147. R01150.–R01734.

1969 R01155.–R01626. R01736.–R02312.01

1970 — R02315.–R03018.

1971 R01629.–R02540. R03021.–R03914.

1973 R02541.–R02688.75 R03920.–R04115.

1975 R02689.–R02850.25 R04126.–R04357.

1976 R02857.–R03714. R04375.01–R05456.50

1978 R03726.–R04059. R05468.10–R05918.

1980 R04064.–R04462. R05955.–R06818.

1981 R04476.–R05479. R06820.–R08118.

1983 R05485.–R05994. —

1990 Sample Person Ques.: R06001.–R07098.Widow Ques.: R07101.–R07871. —

Variable TitlesEvery variable within NLS main file data sets has been assigned a summary title that serves as the verbal representation of that variable throughout the hard copy and electronic documentation system. Variable titles are assigned by CHRR archivists who endeavor, within the limitations described below, to capture the core content of each variable and to incorporate within the title: (1) Key words that facilitate easy identification of comparable variables; (2) universe identifiers that specify the subset of respondents for which each variable is relevant; and (3) for some variables, reference periods that indicate the period of time (e.g., survey year or calendar year) to which these data refer. Universe identifiers and reference periods are discussed below.



Universe Identifiers: If two ostensibly identical variables differ only in that they refer to different universes, the variable title will include a reference to the applicable universe.

Example 1: Universe identifiers are particularly important in the 1990 survey of Older Men. In this survey, “SP” is used to indicate the sample person (the original respondent) and “W” indicates that the widow answered the question. All widow questions are marked as such; if there is no identifier, then the question was addressed to the sample person.‘Year Started Working at Current or Last Job 90’ contains the sample person’s report about the start date of his most recent job.‘Year Started Working at Last Job SP Held, 90 (W)’ contains the widow’s report about the start date of the sample person’s last job before his death.‘Year Started Working at Current or Last Job 90 (W)’ contains the widow’s report about the start date of her most recent job.

Reference Periods: Variable descriptions may include a phrase indicating the time period to which these data refer. The following general conventions apply:

Survey Year: When the variable title includes either the phrase XX INT (81 INT) or the year (e.g., 76) without the year being preceded by the preposition “IN,” this indicates the survey year in which that variable was measured, not necessarily the year to which it applies.

Example 2: ‘Move to Current Residence - Year of (Last) Move, 81’ (Young Men) refers to a residential move described during the 1981 interview.

Example 3: ‘# of Weeks Worked in Past Year, 76’ (Older Men) refers to the weeks worked in the 12 month period preceding the 1976 survey.

Calendar Year: When a date follows a verbal description of a variable and is part of the prepositional phrase “in XX,” the date identifies the calendar year for which the relevant information was collected.

Example 4: ‘Household Record - Family Member # 2: Occupation in 66 (Age 14+) 67’ (Young Men) reports the occupation of the family member during calendar year 1966 as reported during the 1967 interview.



Example 5: ‘Income from Social Security in 70 – R’ (Older Men) refers to payments the respondent received in calendar year 1970 and reported during the 1971 survey.

User Notes: All searches for NLS variables are essentially searches for variable descriptions or titles. Electronic searches of NLS variables via the NLS CD-ROM accessing methods ultimately produce listings of variables by their reference number and variable description or title. The following paragraphs contain cautions regarding possible implications of the search system.

Flexibility in variable title assignment for raw data items is restricted by (1) the actual wording of the question as it appears within the survey instrument; (2) precedent, i.e., how that type of variable has been titled in previous survey years; and (3) the maximum allowable length for variable titles. An attempt is also made to include key phrases in variable titles so that large groups of variables with similar or related subject matter can be easily identified.

Users should be careful not to assume that two variables with the same or similar titles necessarily have the same (1) universe of respondents or (2) coding categories or (3) time reference period. While the universe identifier and reference period conventions discussed above have been utilized, users are urged to consult the questionnaires for skip patterns and exact time periods for a given variable and to factor in the relevant fielding period(s).

Variables with similar content (e.g., information on respondents’ labor force status) may have completely different titles, depending on the type of variable (raw versus created).

Example 6: ‘Employment Status Recode’ (ESR) is the created or reconstructed version of the ‘Activity Most of Survey Week’ raw variable. The ‘Activity’ variable is derived from the first item of the full series of questions used by the Department of Labor (DOL) to obtain employment status; the title reflects questionnaire content. ESR, on the other hand, reflects the procedure used to recode the ‘Activity’ variable. This produces a constructed variable for all NLS respondents based upon responses to the ‘Activity’ question and all other questions used by the DOL to obtain employment status. These other questions serve to qualify and refine employment status beyond the answer to the initial ‘Activity’ question.



Finally, different archivists over a period of three decades performed the task of assigning variable descriptions to data from the NLS cohorts. While every effort has been made to maintain consistency, users may find some differences in variable titles. Two primary sources of variation exist in Original Cohort variable title assignment. The first is systematic error in which identical questions may have the same question wording across the four Original Cohorts but slightly different variable titles. The rule was to make title consistency within a cohort of highest priority. The second variation is attributed to spacing or punctuation errors. The sorting process that produces variable title listings usually places these variables near if not next to the series of interest.

User Notes: There are important differences between the content of telephone and personal interviews. In the late 1960s and early 1970s, most of the interviews were conducted in person, usually at the respondent’s home. There was one attempt at a mail survey in 1968 for the Older Men and the Mature Women; however, the low response rate led to dropping that type of contact. After the first five years, the decision was made to conduct a major survey every five years and two telephone surveys during the five-year span so that problems of recall could be avoided and contact could be maintained with the respondents.

There are several different ways of identifying whether a survey is a personal or telephone interview. Users can: (1) refer to Table 2.4.1 in the “Interview Schedule & Fielding Periods” section of Chapter 2, which depicts the type of interview by survey year, or (2) examine variable titles assigned to questions of similar content. Differences in what appear to be comparable variables reflect variations in the wording of the question or the fact that the reference period for an identically worded question may be different in a personal versus a telephone interview. Questions that refer to the last five years were usually found in a personal (or five-year) interview. This difference means that some questions were only asked in the five-year surveys and some were asked only in the telephone surveys. Users conducting longitudinal analysis need to change their variable creation procedures to account for the differences in data collection between the early years of uninterrupted personal interviews and subsequent survey years when telephone interviews were used.



When analyzing data, users should remember that not all surveys were conducted during the same season of each survey year. Responses to labor force status questions, for example, may differ significantly if fielding occurred during the summer versus winter months. See the discussion of fielding periods in chapter 2 of this guide.

3.3 NLS Codebook SystemAll variables present in the data set are documented via: (1) a cohort-specific codebook, (2) an accompanying codebook supplement, and (3) error updates. This section describes these three primary components of the codebook system and discusses the important types of information found within each.

CodebooksThe codebook is the principal element of the documentation system and contains information intended to be complete and self-explanatory for each variable in a data file. Codebook information can be viewed using the search software or printed from the CD-ROM. This feature enables researchers to customize their documentation for their particular research needs and to select and print information for any or all variables of interest.

Every variable is presented within the documentation as a block of information called a “codeblock.” Each codeblock entry depicts the following important information: a reference number, variable title, coding information, frequency distribution, reference to the questionnaire item or source of the variable, and information on the derivation for created variables. The codeblocks of many variables include special notes containing additional information designed to assist in the accurate use of data from that variable. Each of the above terms is described more completely below.

Codebooks are arranged by reference number. Variables are first grouped according to survey year. Within each survey year, those variables related to the interview (e.g., interview method, interview date, reason for noninterview, sampling weight, etc.) appear first, followed by variables picked up directly from the questionnaire and Information Sheet. In general, created and edited variables appear last, although the created environmental variables are grouped with variables relating to the interview in the early survey years.



User Notes: NLS codebooks are not a substitute for the questionnaires. Although these two pieces of documentation contain similar information, the questionnaires should be used to determine precise universe information.

Codebook Item DescriptionsThe following common types of information for each variable within a codeblock will be discussed in this section: coding information, multiple responses, missing responses, derivations, frequency distribution, questionnaire items (question numbers), universe information, and valid values range. The sample codeblocks in Figures 3.3.1 and 3.3.2 provide visual examples of this information. The sample codeblocks contain similar data items, as both deal with hourly rate of pay variables. However, the first is a question actually asked of respondents during the interview and the second is a created variable; these examples allow users to identify the differences between created variables and questionnaire items. These sample codeblocks are taken from the Older Men data, but the Young Men codeblocks are structured in the same manner.

Figure 3.3.1 Sample Survey Question Codeblock

Reference number

Questionnumber

Frequency distribution

Universe information

Variable title

Survey year

Coding information



Figure 3.3.2 Sample Created Variable CodeblockReference number

Frequency distribution

Universe information

Derivation

Variable title Survey year

Coding information

Valid values range

Coding Information: Each codeblock entry presents the set of legitimate codes that a variable may assume along with a text entry describing the codes. Users should note that coding information in the codeblock for a given variable is not necessarily consistent with the codes found within the questionnaire or for the same variable across years. Use only the codebook coding information for analysis. The following types of code entries occur in NLS codeblocks:

Dichotomous or yes/no variables are uniformly coded “Yes” = 1, “No” = 0. Some dichotomous variables in the 1990 Older Men survey were reformulated to permit this convention to be followed.Discrete (Categorical), as in the case of ‘Highest Grade Attended, 66’ (see Figure 3.3.1).

0 None1 First grade2 Second grade3 Third grade4 Fourth grade

8 Eighth grade9 Ninth grade10Tenth grade11Eleventh grade12Twelfth grade

15Third year college16Fourth year college17Fifth year college18Sixth+ years college-1 Elementary, year



5 Fifth grade6 Sixth grade7 Seventh grade

13First year college14Second year college

unspecified-2 High school, year unspecified-3 College, year unspecified

Continuous (Quantitative), as in the case of ‘Hourly Rate of Pay at Current or Last Job 66 *KEY*’ (see Figure 3.3.2). For continuous variables, the responses are presented in the codebook using a convenient class interval.

000001 thru 999999 rate with two implied decimal places01-499500-9991000-14991500-19992000-2499

2500-29993000-34993500-39994000-44994500-49995000+

Combined Quantitative-Qualitative Variables, i.e., variables which are ostensibly quantitative but which may have several nonquantitative (categorical) responses, utilize positive integers equaling the actual values for the quantitative responses and negative values, beginning with -1, for the qualitative (categorical) responses. For example, the Older Men variable ‘Expected Age of Retirement, 66’ is coded as follows:

45 thru 99 45–99 years-1 age not given-2 will not retire; don’t plan to stop working-3 already retired

Multiple Responses: Response categories to multiple entry questions found in certain Original Cohort job search, discrimination, and health questions have been coded in a geometric progression. More than one response was possible to, for example, the question “What were you doing in the past four weeks to find work?” The response categories to that question were coded as follows:

Checked with public employment agency

1

Checked with private employment agency

2

Checked with employer directly 4Checked with friends or relatives 8Placed or answered ads 16Other method 32

Multiple responses are then coded for each respondent by adding the individual codes, which yields a unique value for each combination. Such multiple entry



variables are identified by an asterisk (*) next to the answer categories in the questionnaire. If a multiple entry has only a few unique combinations, the codebook will specify the exact combinations; those with many combinations need to be unpacked. Methods of unpacking such multiple entry variables are presented in Appendix C at the end of this guide.

Missing Responses: The following conventions were used to treat nonresponse in interviews with the Older and Young Men.

“NA” is the convention used to describe the absence of a valid response where (1) the respondent is not in the applicable universe or (2) the respondent refused to respond or interviewer, coding, transcribing, or data entry error occurred. NA codes are typically treated as missing data.

NA is assigned a value of -128 if valid responses to a question or created variable range from -126 to +127 inclusive.

NA is assigned a value of -999 if valid responses fall outside the range of -126 to +127.

Note: Refusals were also coded as -1 for some income items and other sensitive questions during PAPI interviews; however, -1 has other meanings on other questions, such as ‘Highest Grade Attended, 66’ in Figure 3.3.1. Users should consult the codebook before working with variables that include -1 values.“DK” is the convention used to denote a “don’t know” response; these codes are typically treated as missing data.

DK is assigned a value of -127 if valid responses range from -126 to +127.

DK is assigned a value of -998 if valid responses fall outside the range of -126 to +127.

User Notes: “NEGATIVE” is a convention used in the codebook that provides the frequency of negative responses that are not defined as NA or DK (i.e., missing).

Derivations: The decision rules employed in the creation of constructed variables have been included, whenever possible, in the codebook under the title “DERIVATIONS.” This information is designed to enable researchers to determine whether available constructs are appropriate to their needs. In the ‘Hourly Rate of Pay at Current or Last Job 66 *KEY*’ example (Figure 3.3.2), the derivation describes in detail the questionnaire items used to create the variable. If the



derivation is too lengthy to be included in the codebook, the codeblock instead refers users to the supplemental documentation item that contains variable creation information. In the case of ‘Highest Grade Attended, 66’ (Figure 3.3.1), no derivation is shown because this variable is picked up directly from the questionnaire.

Frequency Distribution: In the case of discrete (categorical) variables, frequency counts are normally shown in the first column to the left of the code categories. In the case of continuous (quantitative) variables, a distribution of the variable is presented using a convenient class interval. The format of these distributions varies. In the case of ‘Hourly Rate of Pay at Current or Last Job 66 *KEY*,’ the frequency count is straightforward. There are twelve categories; the maximum category shown is 5000 and above (since two decimal places are implied, the figure 5000 represents $50.00 and above), for which there is a frequency count of 0.

Questionnaire Item: “Questionnaire item” is a generic term identifying the printed source of data for a given variable. A questionnaire item may be a question, a check item, or an interviewer’s reference item appearing within one of the survey instruments (Table 3.1.1). In ‘Highest Grade Attended, 66’ (Figure 3.3.1), the questionnaire item is 48A.

Universe Information: Universe information for the Original Cohort data sets is printed as separate line items in the codebook for each survey through 1976. Both sample variables present universe information at the bottom of the codeblock; in Figure 3.3.1, for example, 47 respondents do not have information available. Subsequent to 1976, universes can be tracked by referring to the flowchart associated with a particular year’s survey.

Valid Values Range: Depicted below the frequency distribution is information relating to the range of valid values for that particular distribution. “MINIMUM” indicates the smallest recorded value exclusive of “NA” and “DK.” In the case of the created variable example (Figure 3.3.2), ‘Hourly Rate of Pay at Current or Last Job 66 *KEY*,’ this value is 13 with two implied decimal places, or $0.13. “MAXIMUM” indicates the largest recorded value. In the case of the created variable example (Figure 3.3.2), this value is 4619 with two implied decimal places, or $46.19.



Topcoding and Asset Values: To insure respondent confidentiality, income variables exceeding particular limits are truncated each survey year so that values exceeding the upper limits are converted to a set maximum value. These upper limits vary by year and cohort, as do the set maximum values. From 1966 through 1970, upper limit dollar amount variables were converted to set maximum values of 990, 999, 9990, 9999, 999900, or 999999. From 1971 through 1980, upper limit variables were set to maximum values of 50000, and from 1981 to 1983 the set maximum value was 50001. From 1966 through 1980, asset variables exceeding upper limits were truncated to 999999, and beginning in 1981 assets exceeding one million were converted to a set maximum value of 999997. In the 1990 survey of the Older Men, Census also topcoded selected asset items if it considered that release of the absolute value might aid in the identification of a respondent. This topcoding was conducted on a case by case basis with the mean of the top three values substituted for each respondent who reported such amounts.

Codebook SupplementsVariable creation procedures and supplemental coding information are provided within each cohort’s Codebook Supplement. Information provided within these documents is not available in the electronic documentation files on the NLS CD-ROMs. There are separate codebook supplements for the Older Men and Young Men.

Older Men Codebook SupplementAttachment 2: 1960 & 1980 Census of Population Industrial and Occupational Classification Codes provides the occupation-industry coding assignments made by Census Bureau personnel from the verbal descriptions obtained in the interviews. The 1960 Census classifications have been used to code all occupation and industry variables for all survey years. Beginning in 1983, Census started double coding of the current or last job held by the respondent utilizing the 1980 classifications. This attachment also contains a copy of the Duncan Socioeconomic Index, an ordinal prestige scale assigning a rank of 0–97 to each of the three digit 1960 Census occupations.

Attachment 3: Household Record Variables identifies certain relationships and characteristics of family members. This attachment provides, for each member



of the respondent’s household, coding information, variable and reference numbers, as well as frequency distributions for the following types of variables: relationship to respondent; age; school enrollment status; highest grade completed; number of weeks worked and hours worked per week last year; and the occupation code. Users should note that these variables are not present for all survey years for each cohort.

Attachment 5: Employment Status Recodes describes the methodology used by Census to calculate, from the CPS questions that are asked in each NLS survey, each respondent’s employment status. This document provides: (1) definitions of ‘working’, ‘with a job but not at work’, ‘unemployed’, and ‘not in the labor force’; (2) the decision rules used to assign or recode respondents to a particular labor force status; and (3) Census methodology for dealing with exceptions to the rules.

Appendix 1: Table Used for Determining Cutoff Points for Poverty Status Variables

Appendix 2: Additional Instructions to Interviewers for the 1969 SurveyAppendix 3: State Name and State Codes by Census Division Appendix 4: Explanation for Census Processing of Weeks on Layoff (R737. in

1967 and R1248. in 1969)Appendix 5: New Geographic and Environmental Variables 1966–1976Appendix 18: Union Categories—Copy of Coding Instructions for Name of

Union or Employee Association Appendix 19: Derivations for 1981 *KEY* VariablesAppendix 27: Procedures for Calculating the 1990 Weights—R6013.10.,

R6013.20., R7113.10., R7113.20., R7113.30.Appendix 28: Derivations for 1990 *KEY* VariablesAppendix 29: Summary of Major Differences Between the 1990 & Earlier

SurveysAppendix 30: Explanation for the 1990 Created Social Security Variables—

R6743.10.–R6743.30.Appendix 31: Derivations for Collapsed Occupation and Industry Variables—

R7092.–R7098.Appendix 32: Procedures for Coding Cause of Death—Supplemental Death

Certificates



Young Men Codebook SupplementAttachment 2: 1960 & 1980 Census of Population Industry and Occupational Classification Codes provides the occupation-industry coding assignments made by Census Bureau personnel from the verbal descriptions obtained in the interviews. The 1960 Census classifications have been used to code all occupation and industry variables for all survey years. Because the interviews with respondents in the Young Men cohort ceased in 1981, only the 1960 codes were used with this cohort. This attachment also contains a copy of the Duncan Socioeconomic Index, an ordinal prestige scale assigning a rank of 0–97 to each of the three digit 1960 Census occupations.

Attachment 3: Household Record Variables identifies certain relationships and characteristics of family members (early survey years) and household members (later survey years). This attachment provides, for each member of the respondent’s household, coding information, variable and reference numbers, as well as frequency distributions for the following types of variables: relationship to respondent; age; school enrollment status; highest grade completed; number of weeks worked and hours worked per week last year; and the occupation code. Note: Users will not find all of these variables present for all survey years for each cohort.

Attachment 5: Employment Status Recodes describes the methodology used by Census to calculate, from the CPS questions that are asked in each NLS survey, each respondent’s employment status. This document provides: (1) definitions of ‘working’, ‘with a job but not at work’, ‘unemployed’, and ‘not in the labor force’; (2) the decision rules used to assign or recode respondents to a particular labor force status; and (3) Census methodology for dealing with exceptions to the rules.

Appendix 1: Fields of Study in College—Instructions for the Coding SchemeAppendix 2: State Names and State Codes by Census Division Appendix 3: Codes for Enlisted Men and Officers in the MilitaryAppendix 4: Listing of Median Education for Different OccupationsAppendix 5: Listing of Corrections to Employment Status Recode for 1966, 1967,

1968, and 1969Appendix 6: Listing of the Method Used to Create R’s Father and Mother 1966

Variables



Appendix 7: Derivatives for Creation of Dropped Variable Used to Create R2958.Appendix 8: Derivations Used to Create Enrollment Pattern 1966–1969Appendix 9: Determinants of Early Labor Market Success: Appendix AAppendix 10: Determinants of Early Labor Market Success: Appendix BAppendix 11: Determinants of Early Labor Market Success: Appendix CAppendix 12: Determinants of Early Labor Market Success: Method for

Variable Construction Appendix 13: Dual Labor Market Theory: Appendix A (Primary and

Secondary Jobs)Appendix 14: Edits for R2997.Appendix 15: Decision Criteria for Construction of Variables for Dates Last

Attended SchoolAppendix 16: Listing of Corrections Made for 21 Variables (R2315.–R2317.,

R2320.-R2322., R2611.–R2622.)Appendix 17: Work Experience Variables: Time Reference for the 1970

Work History SectionAppendix 18: Union Categories—Coding Instructions for Name of Union or

Employee Association Appendix 19: GED/SVP/Job-Level/Job-Family & 3 Digit Occupation CodeAppendix 20: Derivation for R3836. and Number of Weeks *KEY* Variables

between 1975 and 1976Appendix 21: New Geographic and Environmental Variables 1966–1976Appendix 22: Source for the Job Characteristics IndexAppendix 23: Source for the Job Satisfaction MeasuresAppendix 24: Derivations for R3035., R3829., R3833.Appendix 25: Reason for Preference in Union Certification ElectionsAppendix 26: Derivations for R8114.–R8118.

Error UpdatesPrior to working with a NLS data file, users should make every effort to acquire information on current data and/or documentation errors. A variety of methods are used to notify users of errors in the data files and/or documentation and to provide corrected information. Errors discovered after the release of a data file are distributed in hardcopy form to current purchasers along with the data set. Error notices appear, along with information on how to acquire the corrected data and/or documentation, within NLS News, the quarterly NLS newsletter. NLS News is accessible on-line at www.bls.gov/nls.



3.4 CD-ROM Search FunctionsVariables can be accessed through search and extraction software available on the NLS CD-ROMs. This software provides users with bridging information to the codebook and/or survey instruments. The search indexes and lists described below can be used individually or combined to produce a more refined list of variables. Users should note that, although no distinction is made here, the Older and Young Men data are presented separately on the CD-ROM; users can search for variables and extract data for only one cohort at a time.

Any word search. The “Any Word in Context” function (also called “contextual search”) on the CD-ROM software allows the user to search for and select those variables whose titles contain any single word or combination of words found in the entire documentation database. This function allows users to easily access variables on a variety of topics but is still dependent on the wording of each variable title. For more information on the naming of variables, see the “Variable Titles” discussion in section 3.2 above, especially the User Notes.

Question number list. The CD-ROM contains a searchable list of the question numbers for every variable. By accessing this list, users who know the question number for the item of interest can locate the variable without performing an any word search. Researchers can also browse through the questions for a given questionnaire section.

Reference number list. The disc also contains a searchable list of all reference numbers. By using this list, researchers can locate variables for which they already know the reference number or browse through questions of interest, which are generally arranged in the order in which they were asked during the interview.

Year index. Finally, the CD-ROM includes an option by which users can limit their searches to a single year of interest. Researchers can browse through all the variables relating to a given survey year or can combine the year with an any word or keyword search to locate specific variables of interest.

This section provides only a cursory overview of the search and extraction software on the NLS CD-ROMs. Researchers who need more information should



refer to the quick reference guide in Appendix A of this document and to the CD-ROM User’s Guide distributed with the data.


Chapter 4: Topical Guideto the NLS Men

61

Chapter 4: Topical Guide—Age

4.1 Age

Both the Older and Young Men data sets include information on the respondent’s age as of the initial survey year and date of birth. The initial survey year age variables were provided by the Census Bureau based on information collected during the 1966 household screening. As age inconsistencies were discovered, Census made date of birth information available in the late 1970s. Date of birth and age data were also collected in later surveys during fielding of the “Household Roster” (HHR) section of the questionnaire. Data for reported age and birth dates include a small number of inconsistencies; therefore, attempts to restrict the universe according to age-related variables may have an unwanted result. The cohort-specific User Notes at the end of this section discuss some of the idiosyncratic aspects of these variables.

In addition to the respondent-specific variables discussed in this section, the date of birth and/or age variables for other household members are also available; see the “Household Composition” section of this guide for more information.

Older MenThe Older Men data set contains a respondent age variable (‘Age, 66’) for the initial survey year and two date of birth variables, i.e., ‘Day of R’s Birth,’ ‘Month of R’s Birth,’ and ‘Year of R’s Birth’ for 1966 and a month and year variable for the 1981 survey year. Age at death was calculated by Census at the time of the 1990 interview for 2,660 respondents; data were derived from Census records and month/day/year of death information collected from the widows of respondents. Table 4.1.1 lists reference numbers and source information for these Older Men variables. Table 4.1.2 provides the age of the respondents during the 1966–83 surveys, and Table 4.1.3 presents information about the respondents’ ages at death or the 1990 interview.

Table 4.1.1 Age, Date of Birth & Date of Death Variables: Older Men

Variable 1966 1981 1990

Date of Birth of R R00022.01–R00022.03(Census)

R05051.–R05052.(HHR) —

Age of R R00022. (Census) — —

Age of R at Death — — R07075.–R07077.(Census)



Table 4.1.2 Ages of Interviewed Respondents by Survey Year (Unweighted): Older Men

Age of R on June 30

Survey Year1966 1967 1968 1969 1971 1973 1975 1976 1978 1980 1981 1983

39 140 141 1 142 1 143 1 144 7 1 1 145 278 7 146 386 266 7 1 1 147 374 364 260 648 349 346 359 255 1 149 356 331 346 334 750 355 325 325 325 247 1 151 355 338 315 310 328 7 152 346 338 337 298 313 238 153 346 324 336 322 298 316 6 1 154 316 334 322 319 289 304 232 555 326 301 326 308 301 282 293 210 1 156 309 306 292 308 309 277 285 280 6 157 277 293 293 274 291 288 277 277 198 158 270 263 282 277 296 284 260 261 264 6 1 159 275 261 259 263 263 275 281 246 253 190 660 65 259 254 236 253 278 266 260 245 250 185 161 7 61 248 239 247 246 262 247 232 239 231 662 2 7 60 227 231 241 262 246 240 237 237 17463 2 2 7 57 218 231 235 241 229 219 227 22364 2 2 7 208 212 229 221 223 225 204 23365 2 2 2 52 205 211 218 226 217 211 22166 2 2 7 196 199 190 203 203 205 18867 2 2 48 193 181 202 209 189 19968 2 2 7 176 181 172 188 193 18269 1 2 46 162 165 190 175 18070 2 2 6 43 167 154 183 17971 1 2 6 144 149 140 16072 2 2 2 37 148 142 16573 2 5 128 138 12774 1 2 2 35 120 13275 2 1 6 32 12376 1 2 6 9877 2 1 2 3078 1 1 579 1 2 280 2 182 2

Total Interviewed 5020 4744 4648 4381 4175 3951 3732 3487 3219 3001 2832 2633Not Interviewed — 276 372 639 845 1069 1288 1533 1801 2019 2188 2387Not Available1 13 11 10 7 7 6 4 3 1 1 1 1

Note: This table is based on R00022.01–R00022.03, R05051., and R05052.; the 1981 date of birth variables were used when available. June was imputed for missing months of birth; respondents with missing birth years are in the “not available” category.1 Includes one extreme outlier (age 20 in 1966) who was interviewed in every survey year.



Table 4.1.3 Ages at Death & at Interview by Race (Unweighted): 1990 Older Men

Age at Death Age at 1990 Interview

Age Total Percent Non-black Black Total Percent Non-

black Black

47 2 1 1 148 9 1 3 649 19 1 10 950 9 1 5 451 24 1 14 1052 30 1.1 21 953 30 1.1 23 754 44 1.6 29 1555 45 1.7 27 1856 64 2.4 44 2057 65 2.4 45 2058 73 2.7 46 2759 78 2.9 48 3060 92 3.4 59 3361 89 3.3 61 2862 104 3.9 71 3363 91 3.4 57 3464 109 4.0 75 3465 100 3.7 75 25 1 1 1 —66 114 4.2 69 45 — — — —67 116 4.3 76 40 — — — —68 130 4.8 86 44 3 1 3 —69 133 4.9 76 57 94 4.5 73 2170 119 4.4 82 37 212 10.1 166 4671 136 5.1 87 49 193 9.2 144 4972 114 4.2 77 37 201 9.6 151 5073 117 4.3 80 37 167 8.0 126 4174 92 3.4 69 23 177 8.5 138 3975 85 3.2 57 28 136 6.5 99 3776 99 3.7 67 32 155 7.4 117 3877 80 3.0 57 23 137 6.6 115 2278 64 2.4 47 17 135 6.5 99 3679 57 2.1 44 13 111 5.3 85 2680 42 1.6 34 8 99 4.7 72 2781 41 1.5 32 9 82 3.9 65 1782 21 1 16 5 75 3.6 57 1883 20 1 15 5 68 3.3 59 984 2 1 — 2 33 1.6 25 885 1 1 1 — 2 1 1 186 — — — — 1 1 1 —87 — — — — 1 1 — 189 — — — — 1 1 — 1

Missing 33 1.2 — — 8 1 6 2Total 2693 100 1786 874 2092 100 1603 489

Note: Age at death based on R07075. Age at 1990 interview date based on R00022.01–R00022.03, R05051. and R05052., and R06010.–R06012. The 1981 date of birth variables were used when available; missing birth months were imputed to June and missing birth days were imputed to 15. Missing 1990 interview dates were not imputed.1 Less than one percent.



Young MenA variable for respondent’s age, ‘Age, 66,’ is provided for the initial survey year based on data from the initial household screening. In addition to the 1966 screener birth date information, ‘Day of R’s Birth,’ ‘Month of R’s Birth,’ and ‘Year of R’s Birth,’ birth date information was also collected during administration of the household roster section of the 1976, 1978, and 1981 questionnaires. Table 4.1.4 lists reference numbers and sources for these variables, and Table 4.1.5 provides the age of the respondents during each survey.

Table 4.1.4 Age & Date of Birth Variables: Young Men

Variable 1966 1976 1978 1981

Date of Birth R00022.01–R00022.03(Census)

R05078.–R05080.(HHR)

R05640.–R05642.(HHR)

R07441.–R07443.(HHR)

Age of R R00022. (Census) — — —

Table 4.1.5 Ages of Interviewed Respondents by Survey Year (Unweighted): Young MenAge as of June 30

Survey Year1966 1967 1968 1969 1970 1971 1973 1975 1976 1978 1980 1981

13 4714 554 4615 671 543 4016 703 647 521 3517 595 657 598 487 3618 553 538 590 546 441 3219 427 455 434 506 486 39920 293 364 358 368 488 466 2021 302 262 316 365 393 494 38922 337 275 241 314 397 429 493 1223 354 318 257 228 334 415 533 399 924 307 324 295 256 231 348 458 501 37625 80 286 311 290 251 236 429 524 472 826 74 285 300 295 252 344 450 471 36327 72 267 308 289 226 418 420 446 828 71 264 299 236 342 390 459 351 829 1 69 257 277 231 325 410 431 34530 71 294 226 220 367 436 43931 248 275 209 308 396 42732 67 292 247 210 367 38233 242 271 206 295 36934 65 225 238 206 29635 60 255 201 20136 211 227 19237 57 255 22838 212 24839 53 20840 55

Total Int’d 5225 4790 4318 4033 3993 3987 4014 3977 3695 3538 3438 3398Not Int’d/Avail 11 4361 907 1192 1232 1238 1211 1248 1530 1687 1787 1827

Note: This table is based on R00022.01–R00022.03, R05078.–R05080., R05640.–R05642., R07441., and R07442.; the most recent variables available were used. June was imputed for missing months of birth.



1 Includes one respondent with no birth year available; this respondent was interviewed only in 1966 and 1967.

User Notes: The user is encouraged to carefully examine all age and birth date variables when performing any age-related analysis. Birth data collected at the time of screening may have been provided by a family member, giving rise to possible inconsistencies when comparing a respondent’s reported age with age calculated from date of birth. In cases where age was unknown, interviewers were directed to obtain a “best estimate” of a respondent’s “exact age” at the time of screening and to make corrections later if possible. Furthermore, a respondent may be inconsistent in different interviews in reporting his age. The date of birth inconsistencies are documented in the codebook. Birth date corrections were made in 1981. The birth date corrections should be used carefully and the user is advised to make any additional corrections on a case-by-case basis.

There are varying numbers of out-of-scope cases in the Original Cohorts for two reasons: (1) the birth date variables in a handful of cases are inconsistent with the stated age of the respondent, and (2) some borderline cases that may actually be in scope for the calendar year of the survey are not necessarily in scope at the time the interview took place. CHRR has investigated causes of birth date inconsistencies and has discovered that they arise from birth data originally provided by Census. Unfortunately, these data are generally not recoverable since many of the affected respondents have since attrited. An additional difficulty is that the date of birth for some cases is not reported.

It may be to the user’s advantage to calculate his or her own variable for age based on the reported date of birth. When birth date variables are either unavailable or out of scope, the user may wish to investigate other age-related variables in order to establish age.



Older Men: There are several cases in the Older Men database that are more than one year out of scope when age is calculated from date of birth. Although date of birth was asked again during the 1981 survey, those particular out of scope cases were not surveyed in that year. The user should consult the codebook for case identification codes; other age-related variables may be investigated in order to decide whether to include a particular case in an analysis. In addition, there are 12 cases in the initial survey year with missing birth dates so an attempt to create an age based on date of birth will result, depending on the algorithm used, in missing or possibly even erroneous ages.

Young Men: In the 1981 survey year, there are duplicate date of birth variables for this cohort. One series, R06823., R08068., and R08069. (not depicted on the chart), includes values for individuals interviewed as well as for some who were not interviewed in 1981. It is advisable to use the series of birth date variables R07441.–R07443. for this survey year.

Survey Instruments: Ages and birth dates of respondents were derived from the 1966 household screener. Respondent’s date of birth was collected during select years in the “Household Roster” section of the survey.


Chapter 4: Topical Guide—Alcohol and Cigarette Use

4.2 Alcohol and Cigarette Use

Older MenQuestions on the use of alcohol and cigarettes were asked only in the 1990 Older Men survey. This interview first collected information for both surviving and deceased respondents on the frequency and quantity of alcoholic beverages consumed during the past year. If the respondent had consumed at least 12 drinks during his years as an adult, data were also gathered on the frequency and quantity of the respondent’s lifetime consumption.

The 1990 survey also asked respondents about lifetime cigarette use. Data are available for both current and past users on age when first started smoking regularly and number of cigarettes/packs smoked on a usual day. In addition, the survey asked respondents who had stopped smoking as of the interview date to report their age when they last smoked regularly. Table 4.2.1 reports the ages at which respondents who smoked first began using cigarettes and stopped using cigarettes.

Table 4.2.1 Age at First and Last Use of Cigarettes (Unweighted): 1990 Older Men

Age at last use1

Age at first use1 1–19 20–29 30–39 40–49 50–59 60–69 70+ Don’t

KnowCurrently Smoking Total

1–14 9 17 20 41 91 143 74 8 103 50615–19 13 31 79 120 202 262 145 12 115 97920–24 — 18 35 55 93 95 34 2 47 37925–29 — 2 8 9 28 25 15 1 12 10030–39 — — 2 5 9 14 9 — 4 4340+ — — — — 6 5 4 — 4 19

Don’t Know 2 6 24 40 187 234 170 2592 — 922

Total 24 74 168 270 616 778 451 282 285 2948

Note: Universe is restricted to respondents who reported currently smoking (R06277.) or having ever smoked (R06281., R07196.).

1 Ages reported by living sample members (R06280., R06284., R06285.) or by the widows of deceased respondents (R07197., R07199.). Most “don’t know” responses for age at first use were reported by widows.

2 Includes four cases where the reported age at last use was less than the reported age at first use.

Survey Instruments: The “Health” section of the 1990 Older Men interview and the “Information on Deceased Sample Person” section of the 1990 widows’ questionnaire collected these data.


Chapter 4: Topical Guide—Alcohol and Cigarette Use


Chapter 4: Topical Guide—Aptitude, Achievement & Intelligence Scores

4.3 Aptitude, Achievement & Intelligence Scores

Young MenScores from various aptitude and intelligence tests were collected for respondents in the Young Men cohort during the 1968 survey of high schools. Designed primarily to gather information on the characteristics of the secondary school most recently attended and on respondents’ academic performance, the survey also collected information on the names of the most recent scholastic aptitude or intelligence test (R01711.–R01718.) taken by those respondents who were subjects of the survey. A composite score is available that combines the results from such tests as the Otis/Beta/Gamma, the California Test of Mental Maturity, and the Lorge-Thorndike Intelligence Test, as well as the PSAT, SAT, and ACT college entrance examinations. The composite score is referred to as the ‘IQ Score’ (R01711.) and is described in the appendix to Kohen (1973). Table 4.3.1 lists each test and the number of respondents for whom data are available. Additional information on this survey can be found in the “High School & College Surveys” section of this guide.

Table 4.3.1 Aptitude & Intelligence Tests: Young Men School Survey

Aptitude/Intelligence Test Value in R01712. # of Respondents with Scores

American College Testing Program (ACT/ACTP) 11 44

California Test of Maturity (CTMM/CMM) 2 625

Differential Aptitude Test (DAT) 9 64

Henmon-Nelson Test (HNTMA) 4 216

Iowa Test of Educational Development (ITED) 8 97

Lorge-Thorndike Intelligence Test 3 169

National Merit Scholarship Qualifying Test (NMSQT) 12 20

Otis/Beta/Gamma 1 848

Preliminary & Scholastic Aptitude Tests (PSAT/SAT/CEEB) 7 223

Primary Mental Ability Test (PMA/PMAT) 6 34

School and College Ability Test (SCAT) 10 165

Test of Educational Ability (TEA) 5 42

One assessment, the “Knowledge of the World of Work” scale (R00297.–R00325.), was directly administered to the Young Men in 1966. The first part of this scale asked respondents to pick one of three statements that best described the duties of each of ten jobs commonly held by men. For each of these jobs, the


Chapter 4: Topical Guide—Aptitude, Achievement & Intelligence Scores

respondent then stated how much education was required (i.e., less than a high school diploma, high school diploma, some college, college degree). Finally, the respondent was presented with eight pairs of occupations and asked which of each pair had higher average wages. Several subscores and a total score (R00616.–R00619.) were calculated by survey staff; derivations for these scores are provided in the codebook.

Survey Instruments: The separate instrument called the Young Men and Women School Survey (1968) collected the test score information. The Knowledge of the World of Work scale is located in the “Knowledge of the World of Work” section of the 1966 questionnaire.

Documentation: Appendix 9 in the Young Men Codebook Supplement contains useful background information on the 1968 school survey and details certain variable creation procedures.

User Notes: IQ scores were constructed using scores from the tests available from school records (Kohen 1973). While there may be psychometric problems in constructing an IQ measure from a variety of test forms, these constructed variables assist the user who wishes to construct a unified score from having to repeat the work involved in pooling scores. Similar information is available for respondents in the Young Women cohort.

ReferenceKohen, Andrew I. “Determinants of Early Labor Market Success Among Young

Men: Race, Ability, Quantity and Quality of Schooling.” Ph.D. Dissertation, The Ohio State University, 1973.


Chapter 4: Topical Guide—Class of Worker

4.4 Class of Worker

Older and Young MenIn each survey year, respondents provided data on their class of worker status. Respondents reported whether they (1) worked for a private company or an individual for wages, salary, or commission; (2) were government employees; (3) were self-employed in their own business, professional practice, or farm; or (4) were working without pay in a family business or farm. In addition, level of government data (gathered after 1977) and information on whether a business or professional practice is incorporated (collected at each interview) are included as coding categories in the later survey years. The reference job for these class of worker variables is usually the “current or last job”; however, during the early survey years, the reference job was the “current job.” Definitions for class of worker classifications are listed in Figure 4.4.1.

Each year, survey staff created collapsed versions of the class of worker variables combining the questions described above. These variables distinguish between: (1) wage and salaried workers (including those self-employed respondents who work in an incorporated business); (2) workers self-employed in unincorporated businesses or farms; and (3) those working without pay on family farms or businesses. These collapsed variables are available for all respondents regardless of current employment status; class of worker status for respondents who are unemployed or out of the labor force is derived from the last job reported.

Class of worker data are available not only for the current or last job but also, during select years, for one or more intervening jobs held since the date of the last interview or for dual jobs held during the survey week.

Survey Instruments and Documentation: Questions relating to class of worker can be found in the various “Current Labor Force Status,” “Work Experience,” or “Work History” sections of the Older and Young Men questionnaires. The method of creating the collapsed class of worker variables is provided within each cohort’s codebook.


Chapter 4: Topical Guide—Class of Worker

User Notes: Older and Young Men employment information collected during the early survey years focused on “jobs” while more recent surveys centered on “employers.” Users are urged to carefully consult the survey instruments and to be sensitive to the possibility that persons reporting a new job may still be with their former employer.

Figure 4.4.1 Definitions of CPS Class of Worker Entries

Private Employees are those who work for wages, salary, commission, tips, piece-rates or pay in kind. This applies regardless of the occupation at which the employee worked, whether general manager, file clerk, or porter. Includes persons working for pay for settlement houses, churches, unions, and other private nonprofit organizations.

Federal Government Employees are those who work for any branch of the Federal Government. Includes persons who were elected to paid Federal offices, civilian employees of the Armed Forces, and some members of the National Guard. Also includes employees of international organizations (e.g., United Nations) and employees of foreign governments, such as persons employed by the French Embassy or by the British Joint Services Mission.

State Government Employees are those who work for State governments and include paid State officials (including statewide JTPA [Job Training and Partnership Act] administrators), State police, and employees of State universities and colleges.

Local Government Employees are those who work for cities, towns, counties, and other local areas. Included would be city-owned bus lines, electric power companies, water and sewer service, local JTPA offices, etc. Also includes employees of public elementary and secondary schools.

Self-employed Worker refers to a person working for profit or fees in their own business, shop, office, or farm.

Without Pay refers to a person working without pay on a farm or in a business operated by a related member of the household. Room and board and a cash allowance are not counted as pay for these family workers.

Never Worked refers to a person looking for work who never before held a full-time job lasting two consecutive weeks or more.

Source: Interviewer’s Manual: Current Population Survey. Washington, DC: Department of Commerce, Census Bureau, July 1985.


Chapter 4: Topical Guide—Crime, Delinquency & School Discipline

4.5 Crime, Delinquency & School Discipline

Young MenThe 1968 survey of schools attended by Young Men and Young Women respondents included two questions on whether school records indicated that the respondent had been committed to or was on probation from a correctional institution. Small numbers of respondents from the Young Men cohort were identified: 19 young men had been committed and 13 young men were on probation.

A related question, also asked as part of the school survey, collected information on whether school records indicated that the respondent had ever been expelled or suspended from school. Records of those schools surveyed indicated that 205 Young Men respondents (R01720.) had been expelled or suspended at some point in their schooling.

Survey Instruments: These questions can be found on the separate 1968 School Survey and are located on the data file along with the other high school data collected during 1968.

User Notes: The scope of the Original Cohort surveys is noninstitutionalized individuals; data on incarceration are not available.


Chapter 4: Topical Guide—Discrimination

4.6 Discrimination

Older MenOlder Men answered questions on work-related discrimination in 1971, 1976, and 1980. In all three interviews, respondents reported whether they had experienced discrimination due to age, race, religion, or nationality; sex discrimination was also included in 1976 and 1980. If so, interviews of the Older Men followed up with a question eliciting information on the type(s) of discriminatory practice experienced, e.g., the respondent believed that he was not hired, interviewed, or promoted, or was demoted or laid off because of discrimination. All work-related discrimination questions asked of Older Men referred to the five-year period preceding each interview in which these data were collected.

Young MenThe discrimination questions addressed to Young Men were very similar. In 1971 and 1976, Young Men respondents reported discrimination due to age, race, religion, or nationality; discrimination due to sex was included only in 1976. The questions in the surveys of Young Men referred to a period of two years before the interview date. The follow-up question regarding the type of discriminatory practice experienced was not asked of Young Men respondents.

Survey Instruments: Question numbers by cohort and survey year are as follows:

Older Men: 1971 (Qs 44–46), 1976 (Qs 27–28), 1980 (Qs16–17)Young Men: 1971 (Qs 76–78), 1976 (Q 56)

User Notes: In the 1976 Young Men and 1980 Older Men surveys, the format of the discrimination questions shifted from a single “most important” response to a “mark-all-that-apply” list. These multiple responses have been coded in a geometric progression; users should refer to the “NLS Codebook System” section in chapter 3 and to Appendix C.

ReferenceParnes, Herbert S. Work and Retirement Data: National Longitudinal Surveys of

Middle–Aged and Older Men 1966–1976. Columbus, OH: CHRR, The Ohio State University, 1980.


Chapter 4: Topical Guide—Discrimination


Chapter 4: Topical Guide—Educational Status & Attainment

4.7 Educational Status & Attainment

Older MenBecause the Older Men had already been in the work force for a number of years, schooling was not a major focus for this cohort. The surveys collected basic data on educational attainment, but detailed questions like those asked of the Young Men were not included. In addition to the respondent-specific data, schooling and educational attainment variables are available for select survey years for household members and children of Older Men.

Most information was gathered during the initial survey. Information was collected during 1966 on the highest grade that each respondent had attended and whether that grade had been completed. Follow-up questions, also asked during the 1966 survey, gathered information on whether respondents had been enrolled in a vocational or commercial curriculum during high school, and if so, the primary type of training, e.g., secretarial, bookkeeping, mechanics, woodworking, etc., that they had received. Finally, a created variable reporting each respondent’s highest grade completed is available in 1966.

Some additional data were collected in later surveys of the Older Men. The state in which respondents last attended high school was identified during the 1971 survey. The 1976 interviews gathered information on the highest grade of school that had been completed, the year of graduation from high school, the year last attended college, whether or not a college degree had been received and, if so, the type of highest college degree earned (“associate,” “bachelors,” “masters,” “Ph.D./LL.B./M.D./etc.”). College enrollment during the past year was measured at four survey points: 1973, 1975, 1978, and 1980.

Young MenDue to the fact that schooling, particularly the school-to-work transition process, was a primary focus of the surveys of the Young Men, questions on education were fielded more frequently and the data collection was more comprehensive than with the Older Men. Commonly used educational status and attainment variables available for Young Men respondents are summarized by subject area below. Descriptions of the various standardized test scores available for respondents in the Young Men cohort can be found in the “Aptitude, Achievement, & Intelligence



Scores” section of this guide. Data from the separately administered school survey and the constructed college variables are described in the “High School & College Surveys” section.

Current School Enrollment Status1) Enrollment Status—Is R Currently Enrolled? Whether or not the

respondent was attending regular school at the time of the interview was asked during each survey year.

2) Grade Attending. For those attending regular school, data were collected during each survey year on the specific grade within elementary/high school or the year of college he was attending.

Educational Attainment1) Highest Grade Completed as of XX—Revised. A series of edited

educational attainment variables were created by Center staff for the first nine survey years (1966–76). These are summary variables, available for most respondents, in which each respondent’s record has been longitudinally edited and crosschecked against information gathered during other years. Derivations for most of these revised variables appear within the codebook. Post-1976 highest grade completed variables provide update information for only those respondents attending school since the date of last interview. The “User Notes” section below includes a discussion of these variables.

2) Date of Diploma. Information on the month and year that a high school diploma was received was collected in 1976 for those respondents who had completed high school but who had never been enrolled in college.

3) Ever Attended College. A single question fielded in 1981, the last interview, provides summary information for those respondents not currently attending college (or who reported that they had attended since the last interview) on whether they had ever attended a college or university.

4) Type of College Degree. The information collected in 1966 on type of college degree (“associate,” “bachelors,” “masters,” “doctorate”) was updated during each survey year except 1976 for those respondents who received a degree since the previous interview. The 1976 interview provides information on the highest degree ever received for those in at least the second year of college.

School Experiences1) High School Experiences. The 1966 survey fielded a set of questions

designed to assess each respondent’s overall high school experiences. The high school series included questions on (1) whether or not a respondent participated in extracurricular activities, the number of



hours per week and favorite activity (e.g., sports, music, other clubs, etc.); (2) number of hours per week he spent on homework, where he studied, and if there were distractions to his homework efforts; (3) which high school subjects he liked most/least and the reasons; (4) the kinds of non-school-related activities that took up most of his time (e.g., sports, working for pay, a hobby, etc.); and (5) the respondent’s general attitude toward his high school years.

2) College Experiences. A similar set of 1966 questions assessed the respondent’s overall college experiences. This series collected information on (1) how he felt about his college experience; (2) the field of study liked most/least and the reasons; and (3) for those respondents currently enrolled who would like more education, information was collected on how many years of education he would like to complete, how much education he thought he would actually get, what college he would like to attend, and what field he would like to study.

Comparable sets of questions were asked of respondents in the Young Women cohort. Information collected on the names and locations of up to four colleges was used in construction of the college survey variables; see the “High School & College Surveys” section of this guide for more information.

High School & College Curricula 1) Type of High School Curriculum. Type of high school curriculum, i.e.,

“vocational,” “commercial,” “college preparatory,” or “general,” in which a respondent was enrolled during his last year of high school was collected during the initial 1966 interview. These data were updated during the 1967–71 surveys for respondents enrolled in school during those years.

2) Type of College Curriculum. Information on the type of college curriculum that respondents were pursuing is available for the 1967–71 survey years. The universe for this series is those respondents currently attending college. Coding categories are the same as those used for the ‘Field of Study’ variables described below.

3) Field of Study. Data collected in 1966 on the area of concentration of respondents’ most recent college degree were updated during subsequent interviews for those who received a degree between survey dates. A series of variables were created that summarize the field of study of respondent’s most recent undergraduate/graduate college degree as of the 1966–71 interviews. The 1976 questionnaire asked those respondents enrolled in at least the second year of college for information on the field of study of highest college degree received. The classification system(s) utilized through 1975 included such disciplines as the “Humanities,” “Education,” “Mathematics,” “Business/Commerce,” “Social Science,” “Science,” “Law” while the post-1975 field of study schema were expanded to include such fields as “Computer & Information Sciences,” “Health Professions,” “Public Affairs



and Services,” as well as specific disciplines leading to an associate degree, e.g., “Mechanical & Engineering Technologies,” “Health Services and Paramedical Technologies,” etc.

College Financial Information1) Tuition. Full-time annual tuition amounts were collected during 1966

for the most recent college attended and during 1967–71 and 1976 for the college the respondent was attending that year.

2) Financial Aid. Comprehensive information on financial aid received in connection with college attendance was gathered during all surveys except 1973 and 1975. Data are available on whether a respondent enrolled in college received financial assistance, the type(s) (scholarship, assistantship, loan, GI Bill, etc.), and, for most years, the dollar amount. Users should note that the set of financial aid questions fielded during most survey years as part of the “Assets and Income” section, i.e., whether relatives provided financial aid, which relative(s), and the dollar amount received, did not have as their universe only those respondents attending college.

Type and Location of Schools1) Is School Public? Information on whether the current or last school

attended was public or private was collected during the 1966–71, 1976, and 1981 surveys.

2) Census Division of School Currently or Last Attended. Census division of last high school attended is provided for the 1966 interview. This information was updated during all but the 1978 and 1980 survey years for the school the respondent was currently attending.

3) Comparison of Residence While in High School with Current Residence & with College Location. A series of comparison variables have been created for the 1966–70 survey years that provide information on whether the respondent resided in that year within the same or different county, state, or Census division as that in which his high school or college was located. A second set of variables is present for 1966, 1973, and 1975 that compares the location of, for example, the most recent college attended with other colleges attended. The User Notes below contain a cautionary note on the address information used to construct these variables.

4) Presence & Type of Accredited Two- & Four-Year Colleges in Labor Market of Current Residence. A discrete set of variables was created for the 1966 survey year that provide information on whether there existed within the respondent’s labor market various types of colleges, e.g., two-year public colleges, four-year girls’ colleges, both public and private four-year colleges, etc. The User Notes below contain a cautionary note on the address information used to construct these variables.



Educational Attitudes & Expectations

As discussed under “School Experiences” above, in 1966 a series of questions asked respondents to report their overall attitude toward their high school and college, as well as their favorite and least liked courses. In addition, during five subsequent surveys, Young Men respondents who were still enrolled were asked to report their overall attitude toward school and explain any change relative to the previous survey.

Because the Young Men’s survey was developed to examine the transition from school to work, significant effort was devoted to collecting information on future educational plans. From 1966 to 1976, questions were asked about the respondent’s goal for his completed education (categories range from less than high school to 7 or more years of college), the actual amount of education he expected to receive, and the reason for any change in his educational plans between surveys. For generational comparisons, the respondent’s report of his parents’ goal for his education was collected in 1971 and the respondent’s perception of parental and teacher encouragement for educational goals was gathered in three surveys. Table 4.7.1 provides the reference numbers for these data and indicates universe limitations as applicable.

Table 4.7.1 Young Men: Summary of Information on Educational Plans

Survey YearQuestion 66 67 68 69 70 71 76

Respondent’s educational goal1 R00154. R00584.

R00699. R01103.50

R01227. R01648.50

R01809. R02209.50

R02394. R02968.50

R03116. R03835. R04440.

Comparison of current year’s goal to goal at last interview R00700. R01228. R01810. R03118.

Reason for change in goal2 R00701. R01229. R01811. R02396. R03119.Actual educational level expected R01812. R02395. R03117. R04441.Parents’ goal for resp. at age 14 R03121.

Encouragement from parents/teachers

R02397. R02398. R02400.

R03123. R03124. R03126.

R04443. R04444. R04446.

1 Enrolled respondents only2 Asked only of respondents whose educational goal changed between surveys

In addition to these general aspiration and expectation questions, surveys between 1966 and 1976 asked a number of questions of select universes regarding dates the respondent planned to go back to high school, start college, and return to college. Respondents also reported reasons they planned to end or



continue their education, as well as reasons their specific plans changed since the last interview. Finally, a series of created variables assesses the congruity of the respondent’s educational goal with the educational requirements for his occupational goal.

Survey Instruments & Documentation: The sets of variables described above are found in a variety of questionnaire sections (see below). Appendices within each cohort’s Codebook Supplement present the fields of study classification systems and Census division/state codes.

Older Men: The “Education and Training” section of the 1966 questionnaire, the “Family Background” section of the 1971 instrument, the “Marital History and Other Background” section of the 1976 questionnaire, and the household roster sections of the 1973, 1975, 1978, and 1980 instruments. Young Men: The “Education and Training,” “High School Experiences,” “College Experiences,” and “Educational Goals” sections of the 1966 questionnaire and the “Educational Status” section of subsequent instruments.



User Notes: A commonly asked question concerns the availability of information on highest grade completed. During the initial survey years, the presence of the edited ‘Highest Grade Completed’ variables made the construction of an attainment variable unnecessary. The ‘Highest Grade Completed’ variables were the result of extensive hand-edits; they are, in some ways, a best guess made by examining the complete longitudinal record of each respondent. After the mid-70s, a series of questions was asked during each interview about whether the respondent was currently attending or had attended regular school since the last interview. If the respondent replied in the affirmative, information was gathered on the grade attending and/or completed. These variables are called “update” variables and are available for less than the full universe of respondents. CHRR suggests that researchers needing data on educational attainment for a more complete universe of respondents than those to whom the update questions are administered locate the last summary variable available and use the periodic update information to increment the created variable. In order to simplify the creation process, a global question is fielded every few years to collect information from all respondents on highest grade completed. Problems that arise as the result of data being obtained at multiple survey points will need to be resolved by the individual researcher.

Variables that depend upon address information have been created by Census in an inconsistent manner. The majority of geographic variables were revised in the mid-1970s to correct for known discrepancies in permanent versus temporary address data. However, certain variables, including the ‘Comparison of School Locations with Location of Current Residence’ and ‘Presence and Type of Accredited Two- and Four-Year Colleges in Labor Market of Current Residence,’ were not updated. A more complete discussion can be found in the User Notes in the “Geographic Residence & Environmental Characteristics” section of this guide.


Chapter 4: Topical Guide—Family Background

4.8 Family Background

The Older and Young Men surveys provide researchers with a variety of family background data. The surveys have collected information on three primary topics: parents, siblings (Young Men only), and the respondent’s family background at age 14 or 15.

Older MenParents. In the 1966 interview, each respondent was asked about the birth countries of his parents and grandparents; Table 4.8.1 depicts the results. In addition, the respondent reported the life status of both his parents and his wife’s parents in each personal interview except 1990.

Table 4.8.1 Birth Countries of Older Men Respondents’ Parents & Grandparents

Country Father Mother Maternal Grandfather

Maternal Grandmother

Paternal Grandfather

Paternal Grandmother

US or Canada 3938 3969 3172 3220 3128 3197N or W Europe1 367 319 381 343 402 342C or E Europe2 311 299 33 35 25 22S Europe3 229 208 10 11 13 11Latin America4 29 34 5 4 6 3Other 96 99 5 7 10 13Not available 50 92 1414 1400 1436 1432

Note: This table is based on R00290.–R00295. in the 1966 interview.1 Austria, Belgium, Denmark, England, France, Germany, Iceland, Ireland, Luxembourg, Netherlands, Northern

Ireland, Norway, Scotland, Sweden, Switzerland, Wales.2 Albania, Bulgaria, Czechoslovakia, Estonia, Finland, Hungary, Latvia, Lithuania, Poland, Romania, USSR,

Yugoslavia.3 Andorra, Azores, Gibraltar, Gozo, Greece, Italy, Liechtenstein, Malta, Monaco, Portugal, San Marino, Spain,

Trieste, Vatican City, Europe—Country not specified.4 Mexico, Central America, South America.

Respondent’s background. The 1966 survey asked respondents where they were born and how long they had lived at their current residence. Of the 5,020 respondents interviewed, 267 (5.3 percent) were born outside the United States. The data also include a comparison of the respondent’s birthplace and his residence at the 1966 interview date (e.g., same state; different state, same region; different region; born outside U.S.).

Information was also collected during the initial interview about the living arrangements of Older Men respondents when they were 15 years old, including



with whom the respondent lived and whether the residence was in a large or small city or in the country. In addition, the survey collected the occupation and the highest grade completed of the head of the respondent’s household.

If the respondent had lived with his mother when he was 15, the 1971 interview asked whether she worked for pay at that time. This survey also gathered information about whether a language other than English was spoken in the home when the respondent was 15 years old. If so, the language was recorded; in the data, this information is grouped as follows: Spanish; German, Dutch, Scandinavian (Swedish, Norwegian, etc.); French, Italian, Portuguese; Slavic; and other. Finally, in 1971 the respondent reported the state in which he last attended high school.

Young MenParents. In the 1966 interview, each respondent was asked about the birth countries of his parents and grandparents; Table 4.8.2 depicts the results.

Table 4.8.2 Birth Countries of Young Men Respondents’ Parents & Grandparents

Country1 Father Mother Maternal Grandfather

Maternal Grandmother

Paternal Grandfather

Paternal Grandmother

US or Canada 4976 5008 4301 4381 4228 4320N or W Europe1 58 55 276 238 317 280C or E Europe2 33 34 245 234 219 193S Europe3 48 31 192 160 222 195Latin America4 36 26 64 55 69 65Other 46 36 66 71 77 76Not available 28 35 81 86 93 96

Note: This table is based on R00390.–R00395. in the 1966 interview.1 Country groupings are the same as in Table 4.8.1 above.

In the 1966 interview, the respondent reported the life status of both his parents and his wife’s parents. For all respondents who did not live in their parents’ household, the first four surveys asked about the number of weeks worked by the respondent’s parents, whether they usually worked full- or part-time, and their occupation. This information is available in the “Household Roster” for respondents still residing with their parents.

Siblings. The Young Men provided information about their siblings in two different surveys. The 1966 interview asked respondents how many brothers and



sisters lived outside their home. It also collected the age, highest grade attended, and highest grade completed of the oldest sibling. In 1976, a full collection of information about siblings was included in the survey. Respondents were asked to state the total number of siblings they had and to provide the sex, birth date, and highest grade completed of up to 10 siblings who were not living in their home at the interview date. Information about siblings in the respondent’s household is available in the “Household Roster.”

User Notes: The sampling design for the Original Cohorts produced a number of multiple respondent households. A significant number of Young Men have fathers in the Older Men cohort, mothers in the Mature Women cohort, and siblings in the Young Men and Young Women cohorts. These multiple respondent households provide a great deal of data about a respondent’s parents and siblings in addition to that collected during the regular surveys. For more information, see the “Screening” section in chapter 2 and the “Household Composition” section in this chapter.

Respondent’s background. The 1966 survey asked respondents where they were born and how long they had lived at their current residence. Of the 5,225 respondents interviewed, 124 (2.4 percent) were born outside the United States. As with the Older Men, the data include a comparison of the respondent’s birthplace and his residence at the 1966 interview date.

Information was collected during the initial interview about the living arrangements of respondents when they were 14 years old, including with whom the respondent lived and whether the residence was in a large or small city or in the country. The survey then collected the occupation of the head of the respondent’s household, as well as the highest grade attended and completed by his mother and father. The 1966 survey also asked whether magazines and newspapers were available in the respondent’s home at age 14 and whether anyone in the household had a library card. Finally, the 1971 interview gathered information about whether a language other than English was spoken in the home when the respondent was 15 years old; answers were grouped in the same categories as the Older Men.



Related Variables: In each survey, information is collected on all members of the respondent’s household; see the “Household Composition” section of this guide for details. More information about the respondent’s background is available in the “Race, Ethnicity & Nationality” section of this guide.

Survey Instruments: Information on parents, siblings, and the respondent’s background can be found in the “Family Background” or “Marital History” sections of the questionnaires.


Chapter 4: Topical Guide—Fringe Benefits

4.9 Fringe Benefits

Young MenThis section reviews the fringe benefit data collected for Young Men respondents. Additional information on other work-related benefits can be found in the “Pensions, Social Security & Retirement” section of this guide.

Data on the availability of fringe benefits provided by employers of respondents were collected in 1976 and 1981. Fringe benefits such as: (1) medical/surgical/hospital/dental/vision insurance; (2) life insurance; (3) a retirement pension program; (4) paid sick leave; (5) paid vacation; (6) training/educational opportunities; (7) profit sharing; (8) stock options; and (9) free/discounted meals or merchandise were included in both interviews. In addition, the 1981 survey asked about the availability of flexible work hours. Table 4.9.1 summarizes by survey year the numbers of employed respondents reporting the availability of each type of fringe benefit.

Table 4.9.1 Numbers of Employed Respondents Reporting the Availabilityof Various Types of Fringe Benefits at their Current Job: Young Men

1976 1981Reporting Availability One or More Benefits 2741 2637

No Benefits 262 156

Type of BenefitMedical/Surgical/Hospital/Dental/Vision Insurance 2380 2410

Life Insurance 1932 2024

Retirement Pension Program 1861 1864

Training/Education Opportunities 1362 1484

Profit Sharing 618 654

Stock Options 552 620

Free or Discounted Meals 372 338

Free or Discounted Merchandise 709 605

Paid Sick Leave 2002 1975

Paid Vacation Leave 2440 2324

Flexible Work Hours — 926

Universe: Those respondents who were working at the current survey date within a private company or as a government worker. Table is based on R04561. and R06935.


Chapter 4: Topical Guide—Fringe Benefits

Related Variables: Additional information on eligibility for and receipt of various kinds of retirement benefits and/or on coverage by medical insurance has been collected. Users interested in these sets of questions should refer to the “Income & Assets,” “Health,” or “Retirement” sections of each cohort’s questionnaires.

Survey Instruments: The “Current Labor Force Status,” “Current Labor Force Status and Work History,” or “Employment” sections of the Young Men questionnaires.

User Notes: The universe for the fringe benefit series is restricted to those respondents who have worked since the last interview and who were employed in a private business or as a government worker. The 1981 Young Men data include five respondents who are coded as both with and without benefits; their case ID#s are 601, 1788, 3119, 3206, and 4239.

Multiple entry “mark all that apply” questions typically found in the fringe benefit sections of the Original Cohort questionnaires were coded as geometric progressions during all but the most recent survey years. Program statements to unpack such variables are presented in Appendix C.

ReferenceBureau of Labor Statistics. Work & Family: Changes in Wages and Benefits

Among Young Adults. Report No. 849. Washington, DC: U.S. Department of Labor, July 1993.


Chapter 4: Topical Guide—Geographic Residence & Environmental Characteristics

4.10 Geographic Residence & Environmental Characteristics

Older and Young MenA limited number of geographic variables are available for respondents in the Older and Young Men. Due to Census Bureau confidentiality concerns, such variables provide only broad geographical demarcations of the respondent’s area of residence, e.g., the name of the Census division, whether the residence was located in the South or non-South, and whether the residence was in an SMSA. A series of comparison variables contrast the respondent’s current state/SMSA of residence with those of his birthplace, previous residences, or current job. Specific information on the names of the county, state, or metropolitan statistical area(s) in which respondents reside at given points in time is not available. Finally, characteristics of the respondent’s environment are available from several variables describing the size of the labor force and unemployment rate for the labor market of current residence.

Due to the fact that Census procedures for the geocoding of geographical boundaries were deliberately frozen in the mid-1970s, users are advised to be skeptical about all variables relating to location below the state level except those delineating movement between counties. For more information, see the User Notes below.

Geographic ResidenceSome of the primary sets of geographic variables available for Original Cohort respondents are described below. Table 4.10.1 summarizes the years for which each variable is available for each cohort.

Table 4.10.1 Created Variables for Geographic Residence and Mobilityby Survey Year: Older (O) and Young (Y) Men

Created VariablesSurvey Years

66 67 68 69 70 71 73 75 76 78 80 81 83 90Region of Residence (South/non-South) O,Y O,Y O,Y O,Y Y O,Y O,Y O,Y O,Y O,Y O,Y O,Y O O

Residence Comparison:

State, County O,Y O,Y O,Y O,Y Y O,Y O,Y O,Y O,Y O,Y O,Y O,Y O OSMSA O,Y O,Y O,Y O,Y Y O,Y O,Y O,Y O,Y O,Y O,Y O,Y O O

Size of Labor Market O,Y O,Y O,Y O,Y Y O,Y O,Y O,Y O,Y O,Y O,Y O,Y O

Residence in SMSA O,Y O,Y O,Y O,Y Y O,Y O,Y O,Y O,Y O,Y O,Y O,Y O

Residence Status (Mover/Non-mover) O,Y O,Y O,Y O,Y Y O,Y O,Y O,Y O,Y O,Y O,Y O,Y O O

Unemployment Rate for Labor Market O,Y O,Y Y O,Y Y O,Y O,Y O,Y O,Y O,Y O,Y O,Y O



Birthplace: Birthplace information for each Older and Young Men respondent is available that identifies the birthplace in relation to the respondent’s permanent residence as of the initial survey year. Coding categories differ across cohorts but typically provide a comparison of the respondent’s state/SMSA/region or division of current residence with that of the respondent’s birthplace or identify the birthplace as outside of the United States. Birthplace information is also available for each respondent’s mother, father, and maternal/paternal grandparents; coding categories include U.S. or Canada, Northern or Western Europe, Central or Eastern Europe, Southern Europe, Latin America, or other. The decision rules used to create a nationality variable for each respondent within the four Original Cohorts are discussed within the “Race, Ethnicity, & Nationality” section of this guide.

Region of Residence (Revised): A series of variables are available for Older and Young Men respondents that indicate whether the location of the respondent’s permanent address was “in the South” or “in [one of] the non-South” regions of the United States, e.g., the Northeast, North Central, or West. A listing of states constituting the various Census divisions is provided in each cohort’s Codebook Supplement: Appendix 3 for the Older Men or Appendix 2 for the Young Men. The three divisions comprising the South include the South Atlantic Division, the East South Central Division, and the West South Central Division. Table 4.10.1 above depicts the survey years for which this variable is available. Users should note that two versions, revised and nonrevised, of the ‘Region of Residence’ variables are present. Revised versions should be used whenever available.

Census Division of Current Residence: A series of variables are available for the early years of each cohort that identify the Census division, e.g., New England, Middle Atlantic, Mountain, Pacific, etc., of the respondent’s permanent address. The cohort-specific appendices listed above under ‘Region of Residence’ contain a listing of the nine Census divisions and the states comprising each. This variable for respondents of the 1975 and 1976 Older Men surveys does not provide Census division information but rather the codes “lives in South” and “lives in non-South.”

Residence—SMSA (SMSA Status): A series of revised variables are available for each cohort that identify whether the current residence of a respondent is in the central city of an SMSA, in the balance (not the central city) of an SMSA, or not



in an SMSA. Table 4.10.1 above depicts the survey years for which this variable is available. Two versions of these variables are present within the Young Men data set: ‘Current Residence in SMSA’ and ‘SMSA Status in (YR) (Revised).’ Comparable variables for the Older Men data set are titled ‘Current Residence in SMSA’ and ‘Current Residence in SMSA - Revised.’ The revised version of these variables should be used for those survey years in which it is available. The User Notes section below discusses issues relating to the SMSA classification systems in use by Census.

Residence Status (Mover): A series of revised variables are available which indicate whether a respondent has moved, i.e., reported a permanent address change, since the initial survey year. Residence in the first survey year is coded 1. Code 2 in a subsequent survey year indicates that the respondent has had an address change from the original residence and code 3 indicates that no move occurred. Table 4.10.1 above depicts the survey years for which this variable is available.

Comparisons of Current Residence with Previous State/County/SMSA: This set of variables, available for each survey year, does not provide information on the actual state, county, or SMSA of the respondent’s current residence but rather codes movement of the respondent in relationship to the permanent address reported at the time of the first survey. The respondent’s county, state, and SMSA each are coded 1 for the initial survey year. A code of 2 in a given survey year indicates that the respondent had moved to, for example, a different second county. A subsequent move in year 10 back to the 1966 county would again be coded 1. Appendix 5 in the Older Men Codebook Supplement and Appendix 21 in the Young Men Codebook Supplement provide a further explanation of this coding system along with a select listing of other geographic variables present through the mid-70s for that cohort. The SMSA comparison series was not included in the 1990 Older Men survey for reasons described in the User Notes below. However, two variables, ‘Comparison of 90 Residence with Previous Residence (Movers Only),’ that use a coding system similar to the residence/job comparison variables (see description below) were provided in the 1990 Older Men data set for those respondents or widows of respondents who had moved. Table 4.10.1 above depicts the survey years for which this variable is available.



Comparison of Current Residence & Location of Current Job/Second or Retirement Residence: A set of variables are present for select survey years of each cohort that compare each respondent’s location of current residence with the location of his/her current (or last or longest) job. Coding categories include: same SMSA or county; different SMSA or county, same state; different state, same division; different division; abroad; or other. Several other comparison variables are present for respondents in the Older Men cohort. These include ‘Comparison of XX Residence and Retirement Residence’ for the 1981 and 1983 survey years and ‘Comparison of Location of 2nd Residence and Current Residence 90.’ The User Notes below include a discussion of issues affecting SMSA boundaries.

Second Residence: In 1990, information was collected on whether an Older Men respondent resided in another residence during part of the year. Questions included the specific months of the year the respondent was in residence at that location and the year he first started spending time there.

Geographic Mobility in Retirement: The 1981, 1983, and 1990 surveys of Older Men collected information about residential moves. In 1981 and 1983, respondents who had ever moved since retirement were asked why they moved. In 1990, respondents who had not lived in their current city or county for their entire lives were asked when and why they moved to their current residence.

Type of Property of Residence: Present within the data set of each cohort is a single variable identifying whether the respondent’s property in the original survey year was urban or a farm or non-farm residence with varying acreage and sales.

Type of Area of Residence: A single variable present for the first year of each cohort identifies whether the respondent lived in (1) an urbanized area of a certain size (over 3 million, under 250,000. etc.), (2) an urban place outside an urbanized area of varying population sizes, or (3) a rural area.

Environmental CharacteristicsTwo sets of variables have been created that provide information on characteristics of the labor market in which a respondent resided. The geographical unit used to define “residence” for the revised versions of the following variables was the 1970 Primary Sampling Unit (PSU), a geographical



sampling area made up of one or more contiguous counties or Minor Civil Divisions (MCD).

Residence—Size of Labor Force: A series of revised variables provide information from the 1970 Census of Population on the size of the labor force for the respondent’s current residence. The revised variables are present for the 1966–83 Older Men surveys and all Young Men interviews. The variables based on data from the 1960 Census are present only through the mid-1970s.

Residence—Unemployment Rate for Labor Market: Two series of variables provide data, drawn from the 1970 Census of Population and varying years of the Current Population Surveys, for the unemployment rate of the respondent’s labor market of current residence. These variables are present for all Older Men surveys except 1968 and 1990 and for all Young Men interviews. Unemployment rates were calculated for each CPS PSU by summing the total number of unemployed for the 12-month period and dividing by the total number in the labor force. A combined unemployment rate was computed for PSUs in the same Special Labor Market Areas (combinations of two or more PSUs) and assigned to each PSU within the area.

Survey Instruments & Documentation: These geographic residence variables for the Older and Young Men were, for the most part, created by Census Bureau personnel from the permanent address information available for each respondent. Information on the birthplace of each respondent and of his parents and grandparents was collected during the initial survey year of each cohort; questions can be found in the “Family Background” section of the questionnaires. Information on the location of a current job used to construct the comparison of current residence with location of job was collected as part of the “Current Labor Force Status” sections of the questionnaire.

User Notes: The following paragraphs contain a discussion of issues affecting the availability and quality of geographic data for the Original Cohorts. Researchers are advised to read this information before using geographic variables in analyses.



The amount of geographic information that the Census Bureau has provided to CHRR has always been limited. This was, in part, the trade-off for the richness of data available in all other topical areas. Census felt that the detailed information available for each respondent in combination with the geographic location was sufficient, in some cases, to identify specific respondents. To protect respondent identities and fulfill the promise of anonymity, only gross geographic measures such as South/non-South, size of the labor force from the 1970 Census, and unemployment rate from the 1970 Census and current CPS are consistently released.

As data were analyzed based on respondents’ permanent addresses, some peculiar and inconsistent results were observed. When specifications for the creation of these variables were checked, a problem with the type of address information utilized, permanent versus temporary, was uncovered. It was not clear in all cases exactly which address had been used by Census as the respondent’s permanent address or which respondents had their original data based on address information from the screening as opposed to the first interview. As a result of these problems, the entire series of geographic variables were revised in the mid-1970s.

While in most instances the geographic information from the early surveys will be consistent with that in the revised series, there are sufficient instances when this will not be true. Thus, the revised series should be considered as replacing all earlier geographic information even though the unrevised information has been left on the data sets. Users will find the word “REVISED” appended to the variable titles of most of these variables; the custom of appending REVISED was continued after the mid-1970s revisions to alert users to the fact that the same methodology continued to be utilized to create subsequent years’ variables. Notes that appear within the codeblock of the unrevised variables reference the appendix of the Codebook Supplement that describes the revised variables released at that point in time. It is strongly suggested that this new set of variables be used in any analysis that includes geographic mobility.

After Congress passed the Privacy Act of 1974, Census froze the definitions of NLS geographic variables in an attempt to carry out the spirit of the new law. SMSA codes assigned to the ‘Residence - SMSA Status’ variables were those in effect as



of January 1, 1976 (Office of Management and the Budget). As time passed, these geographic variables became increasingly less useful since the information Census provided was based on definitions that did not correspond to current geographical definitions.

Due to the increasingly inaccurate boundaries and the limitations imposed by the Privacy Act, BLS and CHRR decided to restrict the set of variables that would be created to those that were known to be accurate. For the 1990 Older Men data, the following variables were no longer created: (1) ‘Comparison of Current Residence with Previous SMSA,’ (2) ‘Residence - Size of Labor Force,’ and (3) ‘Residence - Unemployment Rate for Labor Market’ (both Census and CPS versions). Characteristics of the respondent’s local labor market were not released, nor were measures of the geographic proximity of the respondent’s residence to the employer (except what can be approximated by length of travel). Also unavailable is information on whether the location of a respondent’s employer is in an SMSA. Any variables reflecting SMSA status and related comparison variables were discontinued. Retained for continued release were (1) ‘Residence Status (Mover),’ a set of variables that had always been based on permanent address comparisons, and (2) three other variables based on definitions that had remained the same since the inception of the surveys (i.e., ‘Region of Residence [Revised],’ ‘Comparison of Current Residence with Previous State,’ and ‘Comparison of Current Residence with Previous County’). These last two comparison variables never revealed the existing geographic location of the respondent, only his movement into and out of the state and/or county.

ReferencesAdams, Avril and Nestel, Gilbert. “Interregional Migration, Education and Poverty

in the Urban Ghetto: Another Look at Black-White Earnings Differentials.” Review of Economics and Statistics 58,2 (May 1976): 156–66.

Bartel, Ann. “The Migration Decision: What Role Does Job Mobility Play?” American Economic Review 69,5 (December 1979): 775–86.

Maxwell, Nan. “Economic Returns to Migration: Marital Status and Gender Differences.” Social Science Quarterly 68,1 (March 1988): 108–21.

Office of Management and the Budget. Standard Metropolitan Statistical Areas, Revised Edition. Washington, DC: Office of Information and Regulatory Affairs, 1975.



Pitcher, B., Stinner, W., and Toney, M. “Patterns of Migration Propensity for Black and White American Men: Evidence from a Cohort Analysis.” Research on Aging 7,1 (March 1985): 94–120.

Reitzes, D. and Mutran, E. “Location and Well-Being among Retired Men.” The Journals of Gerontology 46,4 (July 1991): S195–203.

Sandell, Steven. “Women and the Economics of Family Migration.” Review of Economics and Statistics 59,4 (November 1977): 406–14.

Tremblay, Carol Horton. “The Impact of School and College Expenditures on the Wages of Southern and Non-Southern Workers.” Journal of Labor Research 7,2 (Spring 1986): 201–11.


Chapter 4: Topical Guide—Health

4.11 Health

This section details the health-related data that have been collected from respondents. In the early survey years, most of the questions focused on the way the respondent’s health affected his work or schooling. As respondents aged, the surveys included more questions about general health, as well as several measures of psychological well-being.

Older MenWork-related variables. Each survey except 1968 asked respondents whether their health limited the kind or amount of work they could do. A supplemental series of questions, fielded during 1966, 1969, 1976, 1981, and 1990, gathered information on the duration of the respondent’s limitation and on whether his health prevented him from working altogether. The 1976 survey determined whether any of the respondent’s reported health problems were the result of an accidental injury and whether the most serious injury occurred on the job. Finally, in 1976 and 1981, retrospective questions asked whether the respondent had ever been prevented from working for 6 or more months due to a health problem and the duration of this limitation.

A series of questions, shown in Figure 4.11.1 below, asked respondents to describe the types of physical activities which their job regularly involved. In several surveys, respondents were also asked to identify specific workplace characteristics that would cause them trouble because of their health.

User Notes: Researchers should be aware that the work limitation questions were not asked in exactly the same format in every survey. In general, more complete information is available in personal interview years.

General physical health. A comprehensive set of health-related variables is available for all respondents. This data collection includes information on the respondent’s health status, perceived health changes over time, and the types of health-related problems and specific health conditions experienced. To provide a general overview of his health, each respondent was asked during the 1966, 1969, and 1978–90 surveys to rate his health as excellent, good, fair, or poor compared to other men his age. Questions were fielded in 1967, 1971, and



1976–83 on whether the respondent considered his health to have remained about the same, improved, or worsened over a set period of time (e.g., the past five years). Self-reported height and weight data are available for respondents interviewed during 1973 or 1990.

Several interviews included question series assessing the respondent’s overall physical condition. These series, and the years when they were included, are described in Figure 4.11.1.

Figure 4.11.1. Older Men Health Question Series

Topic Questions Years

Activities that are difficult

walking using stairs or inclines standing for long periods sitting for long periods stooping/kneeling/crouching lifting/carrying weights < 10 lbs.

lifting/carrying weights > 10 lbs. reaching handling and fingering seeing (even with glasses) hearing dealing with people

19711, 1976, 1981, 1990

Physical problems

pain tiring easily/no energy weakness/lack of strength aches/swelling/sick feeling fainting spells/dizziness

nervousness/tension/anxiety/ depression

shortness of breath/trouble breathing

19711, 1976, 1981, 1990

Working conditions that respondent would have trouble with due to health

fumes/dust/smoke hot places cold places damp places

noise or vibrations confusion or disorder indoors outdoors

19711, 1976, 1981

Activities respondent does regularly at work

walking using stairs or inclines standing for long periods stooping/kneeling/crouching lifting/carrying weights < 10 lbs. lifting/carrying weights > 10 lbs.

reaching handling and fingering reading printed documents hearing special sounds (signals,

directions, etc.) dealing with people

1976

Assistance with daily life can R go outdoors without help can R use public transportation

without help

can R perform personal care without help (bathing, eating, etc.)

frequency R needs help

19711, 1976, 1981, 1990

1 Universe limited to respondents reporting that their health limited the kind/amount of work they could do.

Because the respondents were in their 70s and early 80s at that time, the 1990 survey collected significantly more health information than previous years’ questionnaires. In addition to the questions described elsewhere in this section, the sample person questionnaire asked Older Men who were still living about the topics outlined in Figure 4.11.2.



Figure 4.11.2 Special Health Topics in 1990 Older Men Questionnaire

Topic Questions

Specific health problems in last 12 months

arthritis or rheumatism lung disease (asthma, bronchitis,

emphysema) hypertension heart attack or other heart problem diabetes or high blood sugar cancer or malignant tumor foot problems

stroke broken bones kidney stones/chronic kidney problems back/spine problems paralysis mental illness Alzheimer’s disease

Special equipment needed in last 12 months

glasses/contacts hearing aids cane(s) crutch(es) wheelchair walker leg brace

support stockings artificial limb catheter commode/portable toilet pacemaker knee brace colostomy bag

Use of medical services in last 12 months

number of times in a hospital overnight total number of nights hospitalized number of times in a nursing home total number of days spent in a nursing

home

number of days kept in bed for more than half the day

number of times seen a doctor (other than as part of overnight hospital stay)

Use of community services in last 12 months

special transportation for the elderly meals delivered to home meals at a senior center or program a senior center for any other purpose a homemaker service for the elderly

(cooking, cleaning, etc.)

a service that makes regular phone calls to check on the elderly

visiting nurse home health aide adult day care

Driving has R driven in past 12 months miles driven in past 12 months whether R drives after dark

if no longer driving, when last drove and why stopped driving

Information about institutionalized respondents

when admitted for current stay length of time R will remain in institution where R lived before admission number of times admitted to nursing

home in past two years when admitted for previous stay

who paid for current stay at time of admission

who was paying for current stay at interview date

relationship of R’s emergency contact person

If the respondent was deceased at the time of the 1990 interview, the widow questionnaire gathered information about the last year of the respondent’s life. The widow reported whether and how long the respondent had been ill before his death, the amount of nursing care she provided for the respondent, and the main cause of death. Widows also answered questions, somewhat less detailed than those in the sample person questionnaire, about use of medical services in the year before the respondent’s death. Finally, the widow estimated the total cost of her husband’s medical care in the year before his death, reported the amount paid by the sample person and widow, and described sources of payment for



remaining bills (Medicare, private insurance, employer/union, veteran’s benefits, relatives, etc.).

In conjunction with this survey, up to four causes of death of deceased respondents were recorded from death certificates on file at state vital records departments. These administrative data supplement the cause of death information collected on 2,166 deceased respondents directly from the widows or next-of-kin. Table 4.11.1 compares the causes of death as reported by widows or next-of-kin and as listed on the death certificate.

Table 4.11.1 Cause of Death of Older Men Respondents (1990)

Cause of death Main cause of death reported by widow/proxy

First cause of death listed on death certificate

Heart disease (rheumatic heart disease, heart attack) 893 1037Cancer (neoplasm, malignancy, leukemia) 561 357Stroke 164 145Accident, industrial 6 —Accident, other 56 51Diabetes 55 5Emphysema, chronic bronchitis, asthma 83 183Homicide 10 11Suicide 24 24Other 314 254Total 2166 2067

Note: This table is based on R07185. and R07185.10.

Health insurance. Two surveys collected information on the respondent’s medical insurance. In 1981, the respondent reported whether he was covered by part A of Medicare, part B of Medicare, Medicaid, or any other medical or hospital insurance. The survey also asked if the respondent was a member of a health maintenance organization (HMO) and whether he was eligible for free hospital or medical care due to veteran status. As part of this series, respondents stated whether they had received any hospital or medical care in the past year that was paid for by Medicare, other health insurance, or Medicaid.

The 1990 interview asked sample persons whether they were eligible for Medicare part A and part B, as well as whether they had received any care paid for by Medicaid in past 12 months. If the respondent was covered by a plan other than Medicare, he reported whether it paid hospital expenses and doctor’s bills,



whether it was obtained through an employer or union, and whether the employer or union paid all or part of the premium. If the respondent was not covered by any insurance (other than Medicare), the interview asked him to state the reason. Finally, respondents were asked how they paid for medical expenses not covered by insurance in the past 12 months (savings, current income, debt).

Psychological well-being. A collection of variables related to mental health is available for this cohort. Table 4.11.2 provides reference numbers for the psychological well-being scales and questions described in this section.

Table 4.11.2 Reference Numbers for Older Men Psychological Well-Being Questions

Survey year

Rotter Locus of Control Scale

Bradburn Affect Balance Scale

Pfeiffer Short Portable Mental

StatusCES-D

Depression ScaleAttitudes about

life

1969 R01280.–R01290., R01601.–R01603. — — — —

1971 R02004.–R02014., R02523.00–R02523.02 — — — —

1976 R03103.–R03121., R03707.–R03709. — — — R03123.–R03128.

1978 — — — — R03866.–R03870., R03901., R03902.

1980 — — — — R04255.–R04261.1981 R05008.–R05020. R04865.–R04874. — — R05029.–R05050.1983 — R05700.–R05709. — — R05693.–R05699.

1990 — R06377.–R06386. R07022.–R07031., R07803.–R07812. R06387.–R06396. R06371.–R06376.

In four surveys, the Older Men gave responses to an abbreviated version of Rotter’s Internal-External Control Scale (1966). This scale measures locus of control, with internal control referring to the perception of events as being under personal control and external control meaning that events are perceived as unrelated to one’s own behavior. The abbreviated scale used in the first three administrations included the 11 items from the original 23-item Rotter scale that were the most general and oriented to the adult world of work. The modified scale has been shown to be highly correlated with the original 23-item scale (see Parnes et al. 1974, Appendix to Chapter VI). In 1981, this scale was further reduced to only four items.

Reducing the number of items from 23 to 11 would have resulted in an overall reduction in the range of scores. To avoid this situation, the response format was



modified to four choices rather than the two in the original scale. The respondent was first read a pair of statements representing opposite views and asked which statement was closer to his own opinion. He then stated whether the chosen statement was much closer to his own view or only slightly closer. These answers were combined into one score along a 4-point scale in the data set. The total score was obtained by summing the values of all 11 items, resulting in a range of 11 to 44 (4 to 16 in 1981) in order of increasing external control.

A second scale used in multiple surveys is the Bradburn Affect Balance Scale (Bradburn, 1969). This measure of positive and negative feelings in the weeks before the interview provides an indication of the respondent’s general psychological well-being. The 10 statements used in the Older Men questionnaires asked, for example, whether the respondent felt particularly excited or interested in something in the past few weeks, whether he was so restless he could not sit long in a chair, and whether he felt that things were going his way.

Two scales were used only in the 1990 survey. Addressed only to sample persons (Older Men respondents still living at the time of the interview), the first series used 10 items from the Center for Epidemiological Studies Depression (CES-D) Scale. This scale measures symptoms of depression and discriminates between clinically depressed individuals and others; it is highly correlated with other depression rating scales (see Radloff 1977; Ross and Mirowsky 1989). To provide researchers with an assessment of their cognitive functioning, both sample persons and widows responded to the Short Portable Mental Status Questionnaire (Pfeiffer, 1975). This scale is scored by awarding respondents one point for each correct answer. A score of 2 or less signifies severe cognitive impairment, scores of 3–5 indicate moderate cognitive impairment, and scores of 6 or more indicate mild or no impairment. Pfeiffer cautions that some adjustments need to be made for race and educational level.

Finally, in 1976–90, respondents answered questions about their satisfaction with various aspects of their lives and with life in general. On a 4-point scale, respondents reported their happiness with their housing, the local area in which they lived, their health, their standard of living, and their leisure time activities.



Family member health. Limited information was collected about the health of the respondent’s family members. In most personal interviews, the respondent reported whether his wife was limited in the amount or kind or work or housework she could do and the duration of this limitation. In personal interviews beginning in 1971, the respondent also stated whether his wife’s health problem limited his own work or activities. In 1971, 1976, 1981, and 1990, respondents answered the series of questions about assistance with daily life (see Figure 4.11.1) with respect to their wives. In 1990, interviewed sample persons reported whether their wife was in a long-term care institution; interviewed widows who lived in long-term care institutions (or their proxies) answered the same series of questions as sample persons about the length of and payment for their current stay (see Figure 4.11.2). Finally, the 1981 survey and 1990 sample person questionnaire asked the respondent about his wife’s Medicare coverage and health insurance.

In the 1971 survey only, respondents were asked whether any family members were unable to work or go to school due to a health condition. If so, the respondent then stated whether the family member’s health affected his own employment.

Related Variables: Questions on job satisfaction and other attitudes related to employment can be found in the “Job Satisfaction & Work Attitudes” section of this guide. Attitudes about retirement are discussed in the “Pensions, Social Security & Retirement” section.

Survey Instruments: Health and health insurance questions are located within the “Health” sections of the questionnaires. The CES-D scale items can be found in the “Health” section of the 1990 sample person questionnaire. Components of the Rotter scale can be found in the “Work Attitudes” section of the appropriate instruments. Questions asked of widows are found in the “Information on Deceased Sample Persons” and “Medical Care Prior to Death” sections of the 1990 questionnaire.

Young MenWork-related variables. Each survey except 1967 and 1969 asked respondents whether their health limited the kind or amount of work they could do. In the early years, when some respondents had not yet entered the labor



force, they were also asked whether their health limited their schooling or other activities. In most years, the respondents provided information on the duration of the limitation. The 1976 survey determined whether any of the respondent’s health problems were the result of an accidental injury and whether the most serious injury occurred on the job. Finally, in 1976 and 1981, retrospective questions asked whether the respondent had ever been prevented from working for 6 or more months due to a health problem and the duration of this limitation.

A series of questions, detailed in Figure 4.11.1 above, asked respondents to describe the types of physical activities that their job regularly involved. In several surveys, respondents were also asked to identify specific workplace characteristics that would cause them trouble because of their health.

User Notes: Researchers should be aware that the work limitation questions were not asked in exactly the same format in every survey. In general, more complete information is available in personal interview years.

General physical health. This data collection includes information on the respondent’s health status, perceived health changes over time, and the types of health-related problems and specific health conditions experienced. To provide a general overview of his health, each respondent was asked during the 1971, 1976, and 1981 surveys whether he considered his health to have remained about the same, improved, or worsened over a set period of time (e.g., the past five years). Self-reported height and weight data are available for respondents interviewed during 1973.

Several interviews included question series assessing the respondent’s overall physical condition. The years in which these series were included are provided in Figure 4.11.3.

Figure 4.11.3 Young Men Health Question Series

Topic Questions Years

Activities that are difficult Same as Figure 4.11.1, except seeing, hearing, and dealing with people not asked in 1971 19711, 1976, 1981

Physical problems Same as Figure 4.11.1 19711, 1976, 1981

Working conditions that respondent would have trouble with due to health

Same as Figure 4.11.1 19711, 1976, 1981

Activities respondent does regularly at work Same as Figure 4.11.1 1976



Assistance with daily life Same as Figure 4.11.1 19711, 1976, 1981

1 Universe limited to respondents reporting that their health limited the kind/amount of work they could do.

Psychological well-being. In three surveys, Young Men respondents were administered an abbreviated version of Rotter’s Internal-External Control Scale (1966). For additional information about the content of this scale, researchers should refer to the Older Men discussion above. The Young Men scale is constructed and scored in the same manner as the Older Men version. The reference numbers and years for the Rotter scale administration are as follows:

1968: R01377.–R01387., R01674. (total score)1971: R03526.–R03536., R03898. (total score)1976: R04921.–R04941. (total score not created)

Family member health. Limited information was collected about the health of the respondent’s family members. In most personal interviews, the respondent reported whether his wife was limited in the amount or kind of work or housework she could do and the duration of this limitation. In 1971, 1976, and 1981, the respondent also stated whether his wife’s health problem limited his own work or activities. In 1971, respondents answered the series of questions about assistance with daily life (see Figure 4.11.1) with respect to their wives.

In the 1971 survey only, respondents were asked whether any family members were unable to work or go to school due to a health condition. If so, the respondent then stated whether the family member’s health affected his own employment.

Related Variables: Questions on job satisfaction and other attitudes related to employment can be found in the “Job Satisfaction & Work Attitudes” section of this guide.

Survey Instruments: Health questions are located within the “Health” sections of the surveys. Components of the Rotter scale can be found in the “Work Attitudes” section of the appropriate survey year questionnaires.

ReferencesBradburn, Norman M. The Structure of Psychological Well-Being. Chicago:

Aldine Publishing Co., 1969.



Parnes, Herbert S., Adams, Arvil V.; Andrisani, Paul J.; Kohen, Andrew I.; and Nestel, Gilbert. The Pre-Retirement Years: Five Years in the Work Lives of Middle-aged Men, Vol. 4. Columbus, OH: CHRR, The Ohio State University, 1974.

Pfeiffer, Eric, M.D. “A Short Portable Mental Status Questionnaire for the Assessment of Organic Brain Deficit in Elderly Patients.” Journal of the American Geriatrics Society 23,10 (October 1975): 433–41.

Radloff, Lenore Sawyer. “The CES-D Scale: A Self-Report Depression Scale for Research in the General Population.” Applied Psychological Measurement 1,3 (Summer 1977): 385–401.

Ross, Catherine E. and Mirowsky, John. “Explaining the Social Patterns of Depression: Control and Problem Solving—or Support and Talking?” Journal of Health and Social Behavior 30 (June 1989): 206–9.

Rotter, Julian B. “Generalized Expectancies for Internal vs. External Control of Reinforcements.” Psychological Monographs 80,1 (1966): 1–28.


Chapter 4: Topical Guide—High School & College Surveys

4.12 High School & College Surveys

Young MenThis section describes (1) the separately administered survey that collected information from the high schools attended by respondents in the Young Men cohorts and (2) the set of created variables detailing characteristics of up to three colleges attended by respondents in the Young Men cohorts. Comparable data are available for respondents in the Young Women cohort.

High School SurveyInformation on secondary schools was collected during 1968 by the Census Bureau via a separate school survey mailed directly to the 3,030 schools attended by respondents in the Young Men and Young Women cohorts. After follow-up procedures were conducted to maximize responses, some information is available for approximately 95% of the schools attended by the members of these two cohorts; complete information is available for 75% of the schools (Kohen 1973). Data were collected on (1) characteristics of the schools (type of school, total student enrollment by grade, annual expenditure per pupil, number of books in the school library); (2) characteristics of the school’s teachers and counselors (number of full-time teachers and counselors, annual salary for an inexperienced teacher, presence of a vocational guidance program); and (3) respondents’ performance on various aptitude and intelligence tests as well as their absenteeism and school disciplinary record. Constructed variables including an index of school quality, number of books per pupil, number of students per full-time teacher, number of counselors per 100 students, percent black/Spanish-American student enrollment, and percent black faculty are also available for one or both cohorts. The “Aptitude, Achievement & Intelligence Scores” and “Crime, Delinquency & School Discipline” sections of this guide provide additional information on those subsets of the school survey variables.

Survey Instrument & Documentation: Data were collected using the separate School Survey instrument. The first page of the codebook identifies the reference numbers for these high school variables. A series of appendices within the Young Men Codebook Supplement provides additional information on this survey and some of its constructed variables.



User Notes: The universe for this survey was those respondents who (1) had completed the ninth grade by the time of the 1968 survey and (2) had signed a waiver form permitting Census to collect information from their school records.

College SurveyA series of variables provides information about the colleges attended by respondents in the Young Men and Young Women cohorts during the late 1960s and early 1970s. Data on schooling collected during the regular surveys (e.g., grade attending, when entered this school, names and locations of colleges, highest grade completed) were merged with information detailing the characteristics of each college to form this set of created variables called the “College Survey.”

The following variables were created for each of up to three colleges attended (i.e., the first college attended, the most recent college attended as of 1971, and the college attended for the longest time between the first and most recent college), the last year the respondent attended that college, state identification code for the college’s location, whether the college was private or public, the type of college or university, the highest college degree offered at the institution, the race/sex composition and socioeconomic status of the student body, an index of institutional selectivity, number of books in the library, percentage of faculty with a Ph.D., expenditures per full-time student, ratio of students to faculty, and an index indicating whether the college was “below average,” “average,” or “above average” in six areas of occupational/career orientation.

Survey Instruments & Documentation: Responses to Information Sheet items and data collected from the “Educational Status” sections of the 1966–71 Young Men questionnaires provided the schooling information for each respondent. The first page of the codebook identifies the reference numbers for these college variables. External data sources are identified in the codeblock for each created variable.



User Notes: Respondents who attended fewer than three institutions are coded “NA” for the college attended for the longest time between first and most recent college. For those respondents attending only one institution, characteristics of that institution will be reflected twice, in both the series of variables relevant to the first college attended as well as in those relevant to the most recent college attended.

ReferenceAstin, Alexander. Who Goes Where to College. Chicago, IL: Science Research

Associates, 1965.

Kohen, Andrew. “Determinants of Early Labor Market Success among Young Men: Race, Ability, Quantity and Quality of Schooling.” Ph.D. Dissertation, The Ohio State University, 1973.


Chapter 4: Topical Guide—Household Composition

4.13 Household Composition

This section describes variables related to household and family composition, household identification, linkages between members of multiple respondent households, and household residence. Some familiarity with the following survey instruments which gather information on households is necessary: the Original Cohort Household Record Cards, the Original Cohort “Household Roster,” and the household screeners that were used to select respondents for the various cohorts. The “Survey Instruments and Other Documentation” section of chapter 3 provides detailed descriptions of each of these instruments. This section does not delineate variables pertaining to characteristics or experiences of household members, the presence of partners within the household, or geographical areas of residence. Those interested in information collected specifically on household members should explore the individual topic of interest, e.g., age, sex, educational status, etc. The availability of information on partners is discussed in the “Marriage & Fertility” section of this guide. Finally, those interested in information detailing respondents’ geographic residence, e.g., state, county, or SMSA, should reference the “Geographic Residence & Environmental Characteristics” topical discussion.

Household and Family CompositionDuring each survey year except the 1968 Older Men mail survey, a complete listing of family or household members is available from the “Household Roster” or household enumeration section of the questionnaire. This household section listed household/family members and, for each member, specified information such as that member’s relationship to the respondent, his/her age, school enrollment status, and highest grade completed, and work experience information such as number of weeks worked, hours worked per week, and occupation. In general, household record items were transcribed from the Household Record Cards, documents that were completed before the interview began. As an exception, certain telephone interviews did not update all information on the Household Record Cards; new information was collected directly on the household roster.

User Notes: Information about family or household members collected in the roster section of the questionnaire can be identified by searching for “household record” using the any word search on the CD-ROM.



Older Men. Table 4.13.1 provides a list of the types of information collected in the household roster in each interview. In addition, it depicts whether the roster referred only to members related to the respondent or to all household members. During earlier survey years of the Older Men, only family members (i.e., related household members) were included on the household roster; in more recent years, all household members were listed. However, variable titles were not changed to reflect this difference. Thus, ‘Household Record - Family Member # 5: Relationship to R,’ may actually provide a relationship to the respondent of an unrelated household member. Users should examine the household roster section of the questionnaire to determine whether unrelated household members were listed in a given year.

Table 4.13.1 Older Men Household Roster Questions 1966–90Key: F = Family members, H = Unrelated household members; Numbers indicate

age restrictions

Surv

ey y

ear

Rela

tions

hip

to R

Age/

birth

dat

e (D

oB)

Enro

llmen

t sta

tus

High

est g

rade

at

tend

ed

High

est g

rade

co

mpl

eted

Enro

lled

in c

olle

ge/

univ

. in

past

yea

r?

Wee

ks w

orke

d la

st

year

Usua

l hou

rs/w

eek

work

ed

Occu

patio

n (if

wo

rked

in p

ast

year

)

1966 F F 6-24 F 14 F 14 F 141967 F F age F 6-24 F 6-24 F 6-24 F 14 F 14 F 141969 F F age F 6-24 F 6-24 F 6-24 F 14 F 14 F 141971 F F age F 6-24 F 6-24 F 6-24 F 14 F 14 F 141973 F F age F 14 F 14 F 14 F 141975 F F age F 14 F 14 F 14 F 141976 F F age F 6-24 F 6-24 F 6-24 F 14 F 14 F 141978 F F age F 14 F 14 F 14 F 141980 F F age F 14 F 14 F 14 F 14

1981 F, H F, H DoB F, H 3 F, H 3 F, H 3 F, H 14

F, H 14

F, H 14

1983 F, H F, H both F, H 3 F, H 3 F, H 3 F, H 14

F, H 14

F, H 14

1990 F, H F, H age F, H 3 F, H 14

F, H 14

F, H 14

In addition to the information depicted in the table, the 1981 Older Men household roster included a question asking whether the respondent and each household member were living together in 1976. If not, a follow-up question asked why they



decided to live together (e.g., marriage, health, financial help, share expenses, etc.). These questions were repeated in 1983 with a reference date of 1981.

The 1990 survey design includes more than one household roster. In the sample person questionnaire, Older Men respondents living at home responded to roster questions as in past surveys. In addition to the data in Table 4.13.1, this roster asked whether the respondent and each household member were living together at the date of the last interview and, if not, why they decided to live together. If the Older Men respondent was living in a long-term care institution, a similar but separate roster asked him to report characteristics of the members of his household just before he entered the institution. An additional question inquired whether each member was still living in that household at the interview date. Finally, the respondent reported whether he had lived with each member at the date of his last interview, but the follow-up question was not asked of institutionalized respondents.

The same pattern was repeated in the widow questionnaire, addressed to widows of Older Men respondents. Widows living at home were asked about their current household, including the questions on whether they had lived with each member at their husbands’ last interview date. Widows living in long-term care institutions answered the same questions about their former household that appeared in the sample person questionnaire.

Young Men. Table 4.13.2 lists the information gathered in the various Young Men household rosters in each interview. In all surveys of the Young Men, the actual household roster included only family members. However, in later surveys, some information about unrelated household members was collected on a separate but similar roster. These data are included in the table below and can be located in the data set by searching for the phrase “household members not related to R” in variable titles.



Table 4.13.2 Young Men Household Roster Questions 1966–81Key: F = Family members, H = Unrelated household members; Numbers indicate

age restrictionsSu

rvey

yea

r

Rela

tions

hip

to R

Age/

birth

dat

e (D

oB)

Gend

er

Enro

llmen

t sta

tus

High

est g

rade

at

tend

ed

High

est g

rade

co

mpl

eted

Enro

lled

in c

olle

ge/

univ

. in

past

yea

r?

Wee

ks w

orke

d la

st

year

Usua

l hou

rs/w

eek

work

ed

Occu

patio

n (if

wo

rked

in p

ast

year

)

1966 F F age F 6-24 F F F 14 F 14 F 141967 F F age F 6-24 F 6-24 F 6-24 F 14 F 14 F 141968 F F age F 6-24 F 6-24 F 6-24 F 14 F 14 F 141969 F F age F 6-24 F 6-24 F 6-24 F 14 F 14 F 141970 F F age F 6-24 F 6-24 F 6-24 F 14 F 14 F 141971 F F age F 6-24 F 6-24 F 6-24 F 14 F 14 F 141973 F F age F 14 F 14 F 14 F 141975 F F age F 14 F 14 F 14 F 141976 F F DoB F 3 F 3 F 3 F 14 F 14 F 1419781 F, H F, H both F, H F 3 F 3 F 3 F 14 F 14 F 1419801 F, H F, H both F, H F 3 F 3 F 3 F 14 F 14 F 14

19811 F, H F DoBH age H F 3 F 3 F 3 F 14 F 14 F 14

1 Information for unrelated household members was collected on a separate roster rather than being combined with the regular household roster.

User Notes: Although gender was generally not collected, it can sometimes be inferred from the relationship codes assigned to family and household members. For example, the code for sister identifies a given household member as female. However, some members were assigned a generic code such as “parent—sex not specified”; in these cases, gender cannot be determined.

Users should be aware that relationship codes for family and household members varied across survey years. Figure 4.13.1 below provides information about the relationship codes used in different surveys to classify relationships between the respondent and members of his household.



Figure 4.13.1 Relationship Codes by Survey Year: Older and Young Men

Older Men 1966–78, Young Men 1966–76 Older Men 1980–90, Young Men 1978–81Code Relationship Code Relationship

123456789

10111213

Spouse Child (sex not specified) Son Daughter Parent (sex not specified) Father Mother Parent-in-law Sibling (sex not specified)Sister Brother Grandchild Other relative

1415161718192021222324252627282930

Codes 1-13 as before, plus:Father-in-law Mother-in-law Grandmother or grandparent Grandfather Adopted son or son by marriage Adopted daughter or daughter by marriageOther blood relative-male Other blood relative-female Other in-law relative-male Other in-law relative-female Partner-male Partner-female Boarder-male Boarder-female Other nonrelative-male Other nonrelative-female Other nonrelative (sex not specified)

Household Identification and LinkagesThe sampling design used to select respondents often generated more than one NLS respondent from the same household. More than three-quarters of the respondents from the Young Men cohort and one-third of the respondents from the Older Men cohort shared the same household with at least one other respondent from the same or another cohort at the time the screening was performed (see Tables 4.13.3 and 4.13.4). To facilitate use of this unique aspect of NLS data, constructed variables link respondents sharing the same household at the time of the 1966 screening.



Table 4.13.3 Distribution of Respondents Living within Single &Multiple Respondent Households: The Original Cohorts

Household Type1 Older Men Mature Women Young Men Young WomenSingle Respondent 3353 66.6% 2509 49.4% 1031 19.7% 1018 19.7%Multiple Respondents 1681 33.4 2574 50.6 4194 80.3 4141 80.3

2 Respondent Households 871 17.3 1347 26.5 1997 38.2 1887 36.63 Respondent Households 481 9.6 775 15.2 1206 23.1 1216 23.64 Respondent Households 234 4.6 311 6.1 650 12.4 637 12.35 Respondent Households 71 1.4 115 2.3 264 5.1 300 5.86 Respondent Households 17 0.3 21 0.4 49 0.9 75 1.57 Respondent Households 5 0.1 3 0.1 21 0.4 20 0.48 Respondent Households 1 2 1 2 1 2 5 0.19 Respondent Households 1 2 1 2 6 0.1 1 2

Total Respondents 50343 100% 5083 100% 5225 100% 5159 100%1 Household types for all cohorts are based on data gathered during the household screening. Reference numbers are R00003.–

R00021. (Older Men, Mature Women, and Young Men) and R00003.–R00021.55 (Young Women).2 Less than 0.05%.3 Includes 14 cases later dropped from the public data file.

Table 4.13.4 Distribution of Respondents by Intra- & Inter-Cohort HouseholdsHousehold Type and Cohort(s)1 Older Men2 Mature Women Young Men Young Women Households

Total Respondents 5034 5083 5229 5159 12,382Single Respondent 3353 — — — 3353

— 2509 — — 2509— — 1031 — 1031— — — 1018 1018

Multiple RespondentsIntra-Cohort Respondents3,4 105 — — — 50

— 74 — — 36— — 1697 — 785— — — 1645 743

Inter-Cohort Respondents3,5

OM-MW 574 572 — — 567OM-YM 936 — 1167 — 931OM-YW 843 — — 1069 839MW-YM — 1415 1792 — 1406MW-YW — 1508 — 1957 1502YM-YW — — 2253 2260 1880OM-MW-YM 240 239 306 — 238OM-YM-YW 402 — 513 519 401OM-MW-YW 232 231 — 301 231MW-YM-YW — 618 786 799 614OM-MW-YM-YW 123 122 159 160 122

1 All information on respondents residing in the same household is based on the 1966 screenings. Reference numbers are R00003.–R00021. (Older Men, Mature Women, and Young Men) and R00003.–R00021.55 (Young Women).

2 Includes 14 records later dropped from the public data file.3 Categories are not mutually exclusive. For example, a household containing three Young Men and one Mature Woman would

be included as an intra-cohort Young Men household as well as an inter-cohort Mature Women-Young Men household.4 The number of respondents from households in which at least two respondents from the same cohort resided together at the

time of the 1966 screenings.5 The number of respondents from two or more cohorts who resided in the same household at the time of the 1966 screenings.

Older Men is abbreviated OM, Mature Women is MW, Young Men is YM, and Young Women is YW.



Variables specifying common relationships are available within each Original Cohort data set. These variables provide the identification codes of other respondents originating from the same household by relationship and cohort, e.g., ‘Identification Code of 1st Sister.’ The following relationship linkages are available for the applicable cohorts: spouses, mothers/fathers and children, and siblings. Table 4.13.5 depicts the numbers and types of pairs among Original Cohort members during the initial survey years.

Table 4.13.5 Number & Types of Dominant Pairs Identified during the Initial Survey YearsPairs Young Women Young Men Mature Women

Older Men PairsSpouse Pairs — — 492Parent-Child Pairs 988 1098 —

Mature Women PairsParent-Child Pairs 1848 1671 —

Young Men PairsSpouse Pairs 584 — —Sibling Pairs 1814 902 —

Young Women PairsSibling Pairs 949 — —

Note: This table is based on R00003.50 (Older Men and Mature Women), R00003.01–R00003.52 (Young Men), and R00003.50–R00003.52 and R00021.01–R00021.55 (Young Women).

CHRR staff developed relationship codes based on a Census tape that included the identification numbers of all individuals who shared a household during the screening procedure. The following logic was used in assigning relationship codes: if a 47-year-old man from the Older Men cohort said he had a 38-year-old wife and a 38-year-old woman from the Mature Women cohort with the same household ID said she had a 47-year-old husband, husband-wife relationships were assigned. A one-year difference was allowed between the reported ages; three years of interview information were checked.

Although other types of relationships may have existed, only spouse, sibling, or parent/child relationship codes were assigned. However, identification of other relationship types is possible through use of created variables (R00003. to R00021.) that provide, by cohort, both the ID numbers of other respondents in the household (e.g., ‘Identification Code of 1st Older Male in R’s Household’) and of the household (‘Identification Code of R’s Household’). To determine the nature of other relationships, users can match characteristics of household



members from the first respondent’s information (e.g., the age he claims for a cousin) with characteristics of household members on the second respondent’s household roster (e.g., the age reported for a cousin), as was done for the development of the relationships described above.

Note that phrases such as “Younger Female,” “Older Male,” etc., within the variable titles refer to the cohort—not to the relative age to the respondent. For example, a 14-year-old male has a 17-year-old sister; both are respondents. On his record, she would be called a “Younger Female” because she is in the Young Women cohort.

User Notes: The relationship data were inferred from data on the public data files. CHRR did not have access to detailed information from the Census Bureau (names, etc.) to confirm these linkages. Only “dominant” relationships were considered, as discussed above. While these pairings are believed to be fairly accurate, they and the matching algorithms may have been affected by, for example, misreporting of age in the “Household Roster.”

Although these matches represent unique samples for a number of research topics, users should be aware that they typically include demographically non-representative matches. For example, father-son matches from the Older and Young Men cohorts include fathers who were at least 45 years of age in 1966 and sons who were no older than 24 in 1968.

Once a family relationship was assigned, it was generally considered binding even if the household members lived separately. For instance, if the son of a father/son pair left for college between the screening and the first interview, but their relationship could still be established based on information collected on the father’s “Household Roster” (for anyone away at college), a father/son relationship was assigned. Similarly, if a husband/wife pair was divorced after the initial interview, they would still be linked as spouses. Data from the marital status variables must be used to update the relationship.



The data files for all four Original Cohorts include identification numbers for all other respondents in the household, which can be accessed by searching for the word “Identification.” However, the relationship of the other respondent is not always identified. While identification numbers of spouses in other cohorts are given for all four cohorts, only the Young Men and Young Women files include identification numbers for parent-child pairs. Therefore, for example, a father-son relationship cannot be identified by looking at the Older Men data file; users can only discover that a respondent in the Young Men’s cohort lives in the same household. They must use the Young Men data files to discern whether that pair is a father-son relationship.

Survey Instruments & Documentation: Information on the “Household Roster” is generally transcribed from the Household Record Cards. The “Household Roster” is located within the “Family Members,” “Family Background,” or “Household Members” sections of the questionnaires; information on unrelated household residents was gathered in the “Unrelated Household Member” sections of the Young Men questionnaires. Attachment 3 provides, for each cohort, reference numbers, coding categories, and frequencies for the household record variables.


Chapter 4: Topical Guide—Income

4.14 Income & Assets

Older and Young MenRespondents have been asked numerous questions about their income, assets, and debts over the course of the surveys. While many researchers use income as the primary measure of the economic resources available to a respondent, users can draw a more complete picture of economic well-being by examining both income and wealth. Wealth, which is equal to a respondent’s assets minus his debts, reflects the total financial resources available to the respondent.

IncomeIn every survey year, respondents were asked about their pretax income. Tables 4.14.1 and 4.14.2 present the broad range of income questions asked since 1966, including wages, business and farm income, rental income, interest and dividends, and public assistance support sources. In addition to the in-depth questions about the income of the respondent and his spouse, respondents also provided their estimate for total income of all individuals in the family in some years, while in other years they were asked to estimate the total income of all individuals except their wives and themselves. Finally, respondents have been asked in select years about their ability to get along on their family’s income, with choices ranging from “always have money left over” to “can’t make ends meet.”

User Notes: Researchers should be aware that, in years when the entire survey was shortened, some income sources were dropped altogether; in years when a more in-depth survey was used, the questions were reinserted. Additionally, users should pay careful attention to the wording of income questions to determine exactly whose income is included. In early survey years for the Older Men cohort, respondents were asked about the combined income of themselves and their wives; in later years, Older Men respondents were asked separate questions on how much income they and their wives received from the various sources. In most surveys, Young Men respondents were asked separately about the income of themselves and their wives. However, in the 1973 and 1975 interviews, Young Men respondents were only asked about their own income; there were no questions regarding spouses.



Table 4.14.1 Older Men Income Questions

Survey YearQuestion1 66 67 68 69 71 73 75 76 78 80 81 83 90

Wage & salary * * * * * * * * * * * *Business income * * * * * * * * * * * *Farm income * * * * * * * * * * * *Interest, dividends * * * * * * * * * * * *Rental income * * * * * * *Social Security * * * * * * * * * * * *Pension income * * * * * * * * * * * *Unemployment Compensation * * * * * * * *Workers’ Compensation * * * * * *Disability income * * * * * * *Welfare (AFDC) * * * * * * * * * * * *Food Stamps * * * * * * * * * *Other income * * * * * * * * * * * *Assistance from relatives * * *Total family income * * * * *Ability to get along on income * * * * *

1 All income categories are not asked as separate questions in all years; categories were most often combined in telephone surveys.

Table 4.14.2 Young Men Income Questions

Survey YearQuestion1 66 67 68 69 70 71 73 75 76 78 80 81

Wage & salary * * * * * * * * * * * *Business income * * * * * * * * * * * *Farm income *Interest, dividends * * * * * * * * * * * *SSI * *Unemployment Compensation * * * * * * * * * * * *Workers’ Compensation *Disability income *Welfare (AFDC) * * * * * * * * * * * *Food Stamps * * * * *Inheritance, gifts, prizes *Assistance from relatives * *Total family income * * * * * * * * * * * *Ability to get along on income * * *

1 All income categories are not asked as separate questions in all years; categories were most often combined

in telephone surveys.



AssetsRespondents have periodically been asked a full set of asset and liability questions. Table 4.14.3 depicts the questions pertaining to assets by survey year; note that the 1968 mail survey of Older Men and all telephone surveys are omitted from the table because no asset questions were asked in those surveys. When respondents were asked the full selection of questions, they provided information on the value and mortgage of their home, cash assets, business and farm activity, vehicles, and other debts. In several years, respondents were also asked to rate their overall financial position as better, worse, or about the same as the previous year.

Table 4.14.3 Older and Young Men Asset Questions

Older Men Survey Years Young Men Survey YearsQuestion 66 69 71 76 81 90 66 67 68 69 70 71 76 81

Own home/apartment * * * * * * * * * * *Market value of property * * * * * * * * * * *Amount owed on property * * * * * * * * * * *Have estate/trust *Amount estate/trust *Have money assets * * * * * * * * * * *Amount of money assets * * * * * * * * * * *Have savings bonds * * * * * * * * * * *Amount of savings bonds * * * * * * * * * * *Have stocks/bonds * * * * * * * * * * *Value of stocks/bonds * * * * * * * * * * *Have IRA/Keogh/401k/life insurance *Amount IRA/Keogh/401k/life insurance *People owe you money * * * * * * *Amount owed to you * * * * * * *Own farm/business/real estate * * * * * * * * * *Market value farm/business/real estate * * * * * * * * * *Amount debts farm/business/real estate * * * * * * * * * *Own vehicles * * * * * * * * * * *Owe any money on vehicles * * * * * * * * *Amount owe on vehicles * * * * * * * * *Market value of vehicles * * * * * * *Make/model/year of vehicles * * * * * *Owe money to creditors * * * * * * * * * * *Amount owed to creditors * * * * * * * * * * *Received inheritance/life insurance * *Better/worse financially * * * * * * * * *



User Notes: A number of respondents have wives or parents/children in the other NLS Original Cohorts. If the respondent is part of a multiple respondent household, researchers may be able to compare the respondent’s income and asset information with that provided by other members of his family. (For more information on the possible linkages, users should refer to the “Household Composition” section of this guide.) Using the husband-wife pairs may provide a more complete picture of a respondent’s available resources, while the parent-child pairs provide researchers with information on how income and assets are propagated across generations.

TopcodingTo ensure respondent confidentiality, income variables exceeding particular limits were truncated each survey year so that values exceeding the upper limits were converted to a set maximum value. These upper limits varied by year, as did the set maximum values. From 1966 through 1980, upper limit amounts for both men’s cohorts were set to maximum values of $50,000. For the 1981 surveys and 1983 Older Men interview, the maximum value was set to $50,001 so that researchers could distinguish between topcoded individuals and those reporting $50,000. In the 1990 Older Men survey income amounts exceeding $100,000 were converted to a set maximum value of $100,001.

Asset values were also topcoded. Beginning with the first surveys in 1966, asset variables exceeding upper limits were truncated to $999,999. Beginning in 1976, assets exceeding one million were converted to a set maximum value of $999,994. In 1981 the maximum was increased by three dollars to $999,997. For the 1990 Older Men survey, the Census Bureau topcoded selected asset items if it considered that release of the absolute value might aid in the identification of a respondent. This topcoding was conducted on a case-by-case basis, with the mean of the top three values substituted for each respondent who reported such amounts.

NonresponseOne major concern when asking individuals about their income and wealth is nonresponse bias. While it is outside the scope of this chapter to fully investigate nonresponse bias, this section briefly describes nonresponse in the 1981 survey of both cohorts as an example of the issues raised. There are two primary types



of questions on income and assets (or debts): general questions asking whether the respondent received income from a particular source or owned a particular asset, and specific questions asking about the amount of income or the value of the asset. Factors that are likely to contribute to nonresponse are suspicion, uncertainty, shared responsibility for family finances, and complex financial arrangements. The 1981 survey is examined because it was the last year in which both men’s cohorts were interviewed.

Tables 4.14.4 and 4.14.5 provide information on response rates to questions on income and wealth in the 1981 surveys. Respondents who refused to answer, who responded with “don’t know,” or who are invalid skips are all counted as nonresponses. Both cohorts had high response rates on the receipt and ownership questions—generally around 99 percent. The percentages in the amount column are based only on individuals who reported receiving that type of income or having that type of asset or debt. These amount questions show much lower response rates. For example, business and farm income drops by 20 percent for the Older Men and 13 percent for the Young Men.

Table 4.14.4 Response Rates to 1981 Older Men Income and Asset Questions

Income Category Receive Income from Source Amount1 Asset/Debt Category Own Asset/

Have Debt Amount1

Wages/salaries — 95.9% Money assets 99.6% 74.2%Business 99.5% 79.8% Securities 99.4% 69.1%Interest/dividends 99.5% 69.8% Farm 99.7% 76.7%Social Security 99.6% 84.5% Investment property 99.8% 88.3%Retirement pensions: Business 99.8% 71.2%

Private employer 99.5% 91.4% Vehicles 99.7% 91.4%Military 99.3% 93.3% Primary residence 99.9% 90.7%Federal Gov. 99.4% 92.4% Debts: Mortgage — 95.1%State Gov. 99.4% 79.7% Vehicle debt 99.4% 89.4%Local Gov. 98.5% 85.5%Union 99.4% 86.6%Other pension 99.2% 78.3%

Note: Table is based on R05054., R05059.–R05069., R05072–R05077., R05210.–R05214., R05217., R05218., R05223., R05236., R05237., R05244., R05245., R05268., R05269, R05283.–R05296., R05300., R05301., R05306., and R05307.1 Universe is restricted to individuals who received income from the relevant source or had the relevant asset or debt.



Table 4.14.5 Response Rates to 1981 Young Men Income and Asset Questions

Income Category Receive Income from Source Amount1 Asset/Debt Category Own Asset/

Have Debt Amount1

Wages/salaries/tips --- 94.9% Money assets 99.2% 87.5%Business income 99.1% 87.1% Securities 99.5% 81.5%Unemployment benefits 99.0% 97.8% Savings bonds 99.6% 86.8%Veterans Comp/pension 99.6% 94.7% Farm 99.8% 89.5%Workers’ Compensation 99.5% 98.0% Business 99.8% 85.2%Rental income 96.7% 85.9% Primary residence 99.8% 96.3%Social Security Disability 99.6% 95.2% Debts: Mortgage — 86.7%Farm income 99.6% 86.9% Other debt 99.6% 95.4%

Note: Table is based on R07815., R07819.–R07823., R07826., R07827., R07833.–R07838., R07841., R07842., R07847.–R07851., R07856.–R07861., and R07882.–R07885.1 Universe is restricted to individuals who received income from the relevant source or had the relevant asset or debt.

Created Values and Summary StatisticsCHRR staff have created a small number of summary income and asset variables for both men’s cohorts. The standard variable is ‘Total Family Income’ or ‘Total Net Income of Family.’ This variable is created by adding up all of the individual’s income categories. Should any of the categories be unavailable, the created variable for that year is labeled “not available.” A small number of cases each year have negative income; these individuals have business expenses that are larger than their business and other income. This variable is available in the following survey years:

Older Men: 1966-69, 1971, 1976, 1981, 1990Young Men: 1966-71, 1976, 1978, 1980, 1981

The data sets also include a standard summary variable for wealth entitled ‘Total Family Assets’ or ‘Total Net Family Assets.’ This variable was created by adding up the individual’s housing, savings, bond, IRA, insurance, and business assets and then subtracting mortgages, loans, and other debts. Users are cautioned that a number of respondents have negative net family assets. These variables are present for the following surveys:

Older Men: 1966, 1969, 1971, 1976, 1981, 1990Young Men: 1966, 1970, 1971, 1976, 1981



User Notes: Market value of vehicles was not asked consistently of either cohort. In the years in which these data were collected, automobile wealth was included in the total family assets variable. Researchers should keep this variation in mind when comparing data across interviews.

Additionally, users are cautioned that the 1971 and 1976 Young Men data include two different types of total family asset variables—one set that includes vehicle value and one that excludes it. These variables were created so that users could match the total net asset values from earlier survey years, when automobile wealth was not included in the survey.

Survey Instruments: Each year’s questionnaire has a section on “Income” or “Assets & Income” where the variables described above are located.


Chapter 4: Topical Guide—Industries

4.15 Industries

This section overviews the raw and created industry variables available for the Older and Young Men cohorts. Open-ended questions, e.g., “What kind of business or industry is/was this?”, were included within various sections of the questionnaires during each interview. Verbatim responses to this question were coded by Census personnel using three-digit codes from the 1960, and for select variables, the 1980 classification systems (Census 1960, 1980). Two- and one-digit edited versions of these raw variables are available for each cohort for most survey years. An extensive discussion of Census/CHRR editing and creation procedures that affect the industry variables can be found in the User Notes at the end of this section.

Older MenData were collected during each survey year on the industry of a respondent’s current or last job. The 1967–69, 1976, and 1981 interviews also gathered information on the industry of the longest intervening job held between yearly survey dates or during the past 12 months. The 1971 questionnaire included a detailed work history section that allowed collection of industry information on up to seven intervening jobs. Retrospectives in 1976, 1981, and 1990 collected industry information for the longest job held between 1971 and 1976, 1976 and 1981, and 1983 and 1990. The 1973–76, and 1981 surveys elicited information from those respondents who were unemployed at both the current and previous interviews on the industry of any job held between survey dates. During certain interview years, questions about the industry of a second or dual job, a hypothetical job or business, or a retirement job were fielded. The 1981 and 1990 surveys included questions on the longest job held by the respondent’s wife between 1976 and 1981 or by the respondent’s widow between 1983 and 1990. Information was collected during the 1990 interviews from the widows of deceased respondents on the industry of the last job held by the respondent.

Edited variables from the Occupation & Industry (O & I) Rewrite are present for each survey year that provide one-, two-, or three-digit versions of the raw current/last job variables. The 1960 Census codes were used exclusively up through the 1981 interview to code industry information; beginning in 1983, the current/last job variables were doublecoded with the 1960 and 1980 classification



systems. Several versions of the current/last job variables, e.g., edited and unedited, collapsed and noncollapsed, are also available. See the User Notes below for additional information.

Table 4.15.1 Industry of Current/Last Job by Type of Interview andSample Persons’ Labor Force Status: 1990 Older Men

Industrial Sector TotalSample Person Interview Widow/Proxy

InterviewWorking Not WorkingAgriculture, Forestry, Fisheries 448 64 132 230Mining 49 2 22 22Construction 448 18 176 232Manufacturing 1073 26 430 579Transportation, Communication, Public Utilities 372 13 147 197Wholesale/Retail Trade 552 42 214 275Finance, Insurance, Real Estate 208 30 80 92Business & Repair Services 175 19 68 78Personal Services 118 15 36 64Entertainment & Recreation Services 52 7 18 27Professional & Related Services 382 40 141 180Public Administration 255 14 116 109

Subtotal 4132 290 1580 2085Missing 166 1 38 121Total 42981 291 1618 22062

1 This number reflects the total number of interviews conducted during 1990. Excluded from this table is industry information on 722 respondents for whom neither a respondent nor widow/proxy interview was conducted in 1990. Frequencies in the “Total” column do not equal the sum of the respondent and widow/proxy interview numbers due to the exclusion of 183 interviewed sample persons who do not have valid values on ‘Employment Status Recode,’ the variable used to determine labor force status.

2 For the 2206 respondents whose widow or proxy was interviewed during 1990, the industry code reflects the one provided by the respondent during a previous interview. Information collected during 1990 from a widow/proxy on the industry of the respondent’s last job is available for only a small number of respondents; values are not included in this table.



Table 4.15.2 Industrial Sector of Employed Respondents’ Current Job by Numberof Survey Years with an Industry: NLS of Older Men 1966–83

Industrial Sector Total Ever Employed

Years with Industry1 2 3 4 5 6 7 8 9 10 11 12

Agriculture, Forestry, Fisheries 675 82 73 60 62 42 41 31 36 29 28 38 153Mining 91 23 11 7 4 7 2 6 7 5 3 2 14Construction 839 169 124 68 71 42 51 48 34 26 41 55 110Manufacturing 1780 232 141 133 124 117 125 99 102 77 95 130 405Trans., Comm., Public Utilities 668 106 80 54 43 46 40 37 33 34 22 52 121Wholesale & Retail Trade 1117 233 154 96 93 69 70 69 47 52 49 50 135Finance, Insurance, Real Estate 307 65 32 30 17 24 20 23 13 12 13 12 46Business & Repair Services 374 119 60 46 35 20 21 15 9 9 7 15 18Personal Services 236 64 42 21 11 20 18 13 7 10 6 4 20Entertainment & Recreation Services 94 36 15 8 6 7 6 4 4 0 2 3 3Professional & Related Services 605 108 62 44 54 43 34 34 27 23 28 34 114Public Administration 535 92 56 41 46 25 40 39 24 33 28 25 86

Universe: Respondents who were working or with a job but not at work for whom information on the industry of current employer was available. Industries were coded with the 1960 Census classification system.

Survey Instruments & Documentation: Questions on industry are found in the regularly fielded “Current Labor Force Status,” “Work Experience,” Work History,” and/or “Retrospective Work History” and the special 1966 “Attitudes Towards Work” and 1971 “Plans for the Future” sections of the Older Men questionnaires. Additional industry information was collected using the “Widow’s Work Experience” and “Information on Deceased Sample Persons” sections of the 1990 widow’s questionnaire. Part One and Appendix H of Attachment 2: 1960 & 1980 Census of Population Industrial & Occupational Codes provide listings by industry of the relevant one-, two-, and three-digit codes. Appendix 31 presents a partial derivation for the 1990 collapsed industry variables. Derivations for the collapsed variables from previous years can be found in the documentation for the class of worker variables.

Young MenInformation with which to code the industry of the respondent’s current job or current or last job was collected during each survey year. In addition, all surveys except the 1966, 1978, and 1980 gathered information on either the industry of intervening jobs, the longest intervening job, or the last job held. Five year retrospectives fielded in 1976 and 1981 collected data on the industry of the longest job held during the 1971–76 and 1976–81 periods. Industrial specification of a second or dual job was gathered in 1971, 1976, and 1981. Related questions



fielded during individual survey years included: (1) in 1966, the industry of the job held a year ago, of the job held during the last full year of high school, of the first job held for one month or more after the respondent stopped attending school full-time, and of a (hypothetical) job or own business that the respondent would start should he lose his current job; and (2) in 1971 and 1976, the industry of an alternative job that those respondents who reported job-shopping while remaining employed with the same firm indicated that they could have had and/or had been offered. Edited variables from the Occupation & Industry (O & I) Rewrite are present for each survey year that provide one-, two-, and three-digit versions of many of these raw variables. The 1960 Census classification system is used exclusively to code industry information for this cohort. See the “User Notes” section below for additional information. Finally, a variable entitled ‘Industrial Diversification Score for Labor Market of Current Residence’ was constructed for the 1966–70 survey years that provides an index of the degree to which the industrial composition of the respondent’s local area resembles that of the national economy.

Survey Instruments & Documentation: The regularly fielded “Current Labor Force Status,” “Work Experience & Attitudes,” and “Retrospective Work History,” and the special 1966 “Attitudes toward Work” sections of the Young Men questionnaires contain questions on industry. Part One and Appendix H of Attachment 2: 1960 & 1980 Census of Population Industrial & Occupational Codes provide one-, two-, and three-digit codes.



Table 4.15.3 Industrial Sector of Civilian Respondents’ Current/Last Jobby Survey Year: Young Men 1966–81

Industrial SectorSurvey Year

66 67 68 69 70 71 73 75 76 78 80 81All Respondents 4574 4415 4167 3893 3878 3935 3968 3949 3643 3493 3402 3296Agriculture, Forestry, Fisheries 571 481 340 243 213 214 178 163 154 146 153 158Mining 22 22 29 33 40 35 46 73 55 56 64 64Construction 380 363 405 401 400 422 478 446 412 359 367 347Manufacturing 1140 1092 1132 1185 1167 1129 1218 1168 1052 1005 959 898Transportation, Communications, &

Public Utilities 190 209 249 246 258 280 304 327 292 287 311 300

Wholesale & Retail Trade 1130 1080 993 854 757 790 671 627 590 562 469 482Finance, Insurance, Real Estate 79 85 87 89 121 139 152 158 152 150 148 148Business & Repair Services 178 143 137 133 142 130 156 148 159 137 177 175Personal Services 240 200 117 77 81 74 66 56 47 39 44 40Entertainment & Recreation Services 131 124 91 60 55 47 47 42 39 37 28 35Professional & Related Services 382 463 437 421 487 490 445 481 456 460 442 425Public Administration 131 153 150 151 157 185 207 260 240 255 240 224

Universe: Respondents both working and not working during the survey week for whom an industry code from the 1960 Census classification system for their current or last job was available.

Table 4.15.4 Industrial Sector of Civilian Employed Respondents’ Current Jobby Number of Survey Years with an Industry: Young Men 1966–81

Industrial SectorTotal Ever Employed

Years with Industry1 2 3 4 5 6 7 8 9 10 11 12

Agriculture, Forestry, Fisheries 913 341 199 118 57 35 34 23 23 21 13 19 30Mining 182 80 28 19 15 11 8 7 6 2 3 0 3Construction 1558 653 296 159 100 87 66 41 51 29 37 15 24Manufacturing 3077 796 511 342 270 225 199 155 147 117 110 88 117Transportation, Communications, &

Public Utilities 1062 443 195 125 58 59 54 31 23 25 17 17 15

Wholesale & Retail Trade 2729 906 577 366 244 177 123 83 64 63 54 36 36Finance, Insurance, Real Estate 499 216 91 44 26 31 30 22 13 10 9 3 4Business & Repair Services 860 489 156 82 51 33 16 13 7 3 3 3 4Personal Services 552 342 128 38 11 3 10 3 9 2 1 3 2Entertainment & Recreation Services 384 230 87 29 13 11 5 3 1 4 0 1 0Professional & Related Services 1472 485 263 177 124 96 67 54 63 45 35 31 32Public Administration 710 274 122 69 55 50 53 21 16 15 9 10 16

Universe: Respondents whose activity most of the survey week was working or with a job but not at work who reported information on their employer from which an industry code from the 1960 Census classification was assigned. Excluded are industry codes for jobs reported in the separate series of questions about jobs held during high school and the first job held after high school.



User Notes: For both the Older and Young Men, substantive differences exist between a number of similarly titled occupation, industry, and class of worker variables present in the data files. One set of raw variables relating to the respondent’s current job is derived from responses to questions found within the “CPS” section of each questionnaire. Additional versions of this set of variables are created using the two different procedures described below.

(1) An Occupation & Industry (O & I) Rewrite creates a set of seven “backfilled” or summary variables that enable researchers to identify the last occupation, industry, or class of worker status of all respondents who were interviewed in a given year whether or not they were currently working. Values utilized are those from the job in which the respondent was employed the week before the interview or “backfilled” values from the job that was current at the last time the respondent reported employment. Although the industry associated with an intervening job might technically be a respondent’s most recent industry affiliation, the O & I program is not designed to pick up information from such jobs. All O & I variables are classified utilizing the 1960 Census codes. Titles for this set of O & I Rewrite variables appear below (Table 4.15.5)

Table 4.15.5 Occupation & Industry Variables from the O & I Rewrite

Variable Title Version Question #Class of Worker at Current or Last Job CollapsedOccupation of Current or Last Job 3-digitOccupation of Current or Last Job Duncan Index [Always Blank—Occupation of Current or Last Job 1-digit Created Variables]Industry of Current or Last Job 3-digitIndustry of Current or Last Job 2-digitIndustry of Current or Last Job 1-digit

O & I Rewrite variables can be differentiated from non-backfilled variables by (1) the absence of a question number in the documentation that identifies the source of the variable; or (2) an assignment of a reference number that places an O & I variable among the created variables series appearing at the end of a given survey year’s variables. In order to assist researchers identify these variables, an attempt has been made to append the word “collapsed” to the end of the O & I variable titles of select cohorts. This convention has been applied as follows: (1) all O & I ‘Class of Worker’ variables in all four cohorts for all survey years have the word



“collapsed” appended to the variable titles and (2) the O & I occupation and industry variables from the 1990 Older Men surveys are identified with the word “collapsed.” O & I Rewrite variables for the Young Men can only be identified using the question/reference number assignment conventions discussed above. Users unfamiliar with NLS assignment conventions should refer to chapter 3 of this guide.

It is not clear why the O & I Rewrite variable titles are identical to those of the occupation/industry/class of worker variables. It may have been that these O & I rewrite variables were originally private variables intended only for use by in-house CHRR researchers and subsequently released to the public.

(2) An editing procedure begun by Census in the 1980s cleans items from the “CPS” section of the questionnaire in order to create the ‘Employment Status Recode’ variables. The ESR variables were originally generated by Census with no cleaning or editing of the items from the “Current Labor Force Status - CPS” sections of the questionnaire. In the mid-1980s, reoccurring problems with the program that created ESR forced Census to create edited “CPS” items. Both unedited and edited versions of these items are sent to CHRR and released to the public. Edited variables are identified with either the word “EDITED” or the abbreviations “EDT” or “E” appended to the variable title. Edited versions of these variables will have fewer cases than the unedited versions. When looking at patterns over time, users may wish to use the set of unedited versions.

ReferencesBlackburn, McKinley L. and Neumark, David. “Unobserved Ability, Efficiency

Wages, and Interindustry Wage Differentials.” Quarterly Journal of Economics 107,4 (November 1992): 1421–36.

D’Amico, Ronald. “Career Paths and Career Origins: The Effect of First Job Industry on the Attainments of Mature Men.” Columbus, OH: CHRR, The Ohio State University, 1981.

D’Amico, Ronald. “A Quantitative Procedure for the Assignment of Industries to Capital Sector Types.” Columbus, OH: CHRR, The Ohio State University, 1981.

D’Amico, Ronald and Daymont, Thomas. “Industrial Organization, Economic Conditions, and the Labor Market Success of Young Men.” In: Market Defenses: Early Work Decisions of Today’s Middle-Aged Men, S.M. Hills, ed. Columbus, OH: CHRR, The Ohio State University, 1983.



Jacobs, Jerry A. “Industrial Sector and Career Mobility Reconsidered.” American Sociological Review 48,3 (June 1983): 415–21.

Shapiro, David and Hills, Stephen M. “Adjusting to Recession: Labor Market Dynamics in the Construction, Automobile, and Steel Industries.” In: The Changing Labor Market: A Longitudinal Study of Young Men, S.M. Hills, ed. Lexington, MA: Lexington Books, 1986.

U.S. Census Bureau. 1960 Census of Population Alphabetical Index of Occupations and Industries (Revised Edition). Washington, DC: U.S. Government Printing Office, 1960.

U.S. Census Bureau. 1980 Census of Population Classified Index of Industries and Occupations. Washington, DC: U.S. Government Printing Office, 1980.


Chapter 4: Topical Guide—Job Characteristics

4.16 Job Characteristics

Young MenThe variables discussed in this section are limited to the special 1978 data collection for the Young Men cohort. This survey included a series of questions on characteristics of the respondents’ current job, e.g., the amount of variety, autonomy, opportunity to deal with people or develop friendships or complete tasks, as well as the amount of significance they attributed to their job and the amount of performance feedback received. Items for this scale, the Job Characteristics Index (JCI), was developed by Sims, Szilagyi, and Keller and is an extension of the work first begun by Turner and Lawrence in 1965. The JCI was preceded by an instrument developed by Hackman and Oldham known as the Job Diagnostic Survey (JDS), dimensions of which are also incorporated in the JCI, although in a simpler format. Comparisons of the JCI and JDS by Dunham et al. have shown that both scales tend to collapse to a one-dimensional scale measuring job-complexity. Therefore, the JCI was shortened by selecting one scale item that loaded strongly on each of the dimensions of job complexity shown to be important in earlier research. In their 1976 article, Sims et al. reported the necessary factor analysis scores used to obtain the abbreviated scale.

Question and reference numbers for the seven items that comprise the shortened JCI scale are as follows:

R05542.–R05548. (Questions 12A–12G)R05570.–R05575. (Questions 16A–16F)

Survey Instruments: These questions are found within the “Current Labor Force Status or CPS” section of the Young Men questionnaire.

ReferencesDunham, Randall B.; Aldag, Ramon, and Brief, Arthur P. “Dimensionality of Task

Design as Measured by the Job Diagnostic Survey.” Academy of Management Journal 20, 2 (June 1977): 209–23.

Hackman, J.R. and Oldham, J.R. “Development of the Job Diagnostic Survey.” Journal of Applied Psychology 60 (1975): 159–70.

Hills, Stephen M. “Attitudes of Union and Nonunion Male Workers toward Union Representation.” Industrial & Labor Relations Review 38,2 (January 1985): 179–94.


Chapter 4: Topical Guide—Job Characteristics

Pierce, Jon L. and Dunham, Randall B. “The Measurement of Perceived Job Characteristics: The Diagnostic Survey vs. the Job Characteristics Inventory.” Academy of Management Journal 21,1 (March 1978): 123–28.

Sims, Henry R.; Szilagyi, Andrew; and Keller, Robert. “The Measurement of Job Characteristics.” Academy of Management Journal 26,2 (June 1976): 195–212.

Turner, A.N. and Lawrence, P.R. Industrial Jobs and the Workers: An Investigation of Responses to Task Attributes. Boston: Harvard University Press, 1965.


Chapter 4: Topical Guide—Job Satisfaction & Work Attitudes

4.17 Job Satisfaction & Work Attitudes

Older and Young MenA variety of questions were asked of respondents in the men’s cohorts that assessed both attitudes toward their own employment and attitudes about work in general.

Global job satisfaction. During most survey years, respondents described how they felt about either their current job or their current/last job. The basic question in this series asked respondents to rate their general attitude on a scale from “like it very much” to “dislike it very much.” In some surveys, the general question was followed by an open-ended request for a description of the aspects of his job the respondent liked most and least. These responses were coded and included in the data set. Finally, data comparing respondents’ attitude toward their current job with their attitude toward their job in a previous survey year were collected during certain early surveys of each cohort. Tables 4.17.1 and 4.17.2 provide information about the years in which these questions were asked and the reference numbers for each item.

Facet-specific job satisfaction scale. During the 1978 and 1981 surveys of Young Men, employed respondents (wage and salary workers or self-employed respondents) were asked a series of detailed questions relating to specific aspects of their jobs. On a scale from “very true” to “not at all true,” respondents rated a series of descriptive statements about the pay, working conditions, chances for promotion, job security, competency of their supervisor, and friendliness of their coworkers. These variables can be located by searching for the phrase “Job Satisfaction Index”; reference numbers are not included in Table 4.17.2.

User Notes: The job satisfaction questions were sometimes asked multiple times in a single survey, with each question addressed to a different universe of respondents. Similarly, the facet-specific scale was asked separately of self-employed Young Men and Young Men employed as wage and salary workers, with slight differences in the items included. To obtain a picture of job satisfaction for all respondents, it will be necessary to combine these items.



For the Young Men, a job satisfaction index can be constructed by combining (a) the global job satisfaction measures, (b) select items from the facet-specific job satisfaction ratings, and (c) responses to a question on whether the respondent would stay in his current job if he were free to take any job.

Motivation and commitment to work. Both Older and Young Men respondents answered questions assessing their motivation for working and their commitment to their jobs. The motivation question asked the respondent whether liking his work or receiving high wages was more important. The commitment to work questions first asked whether the respondent would continue to work if he obtained enough money to live comfortably; open-ended follow-up questions asked the respondent to provide a reason for his answer. Finally, respondents in both cohorts were asked whether they would prefer to work more hours for more money, work fewer hours for less money, or maintain their current hours and compensation. The “other work attitudes” columns in Tables 4.17.1 and 4.17.2 below list the years in which these questions were asked of each cohort and the reference numbers for the various items.

Retrospective evaluation of work. Respondents were occasionally asked to look back over a period of time and assess changes in their employment. The first set of questions asked both Older and Young Men whether they felt that they had progressed, held their own, or moved backward in the past 5 years. If they had progressed or moved backward, respondents provided additional information about the specific ways in which that movement had occurred (e.g., changes in wages; hours; fringe benefits; status, level, or responsibility; job security). A second series asked whether the respondent felt that the pressures of his job had changed in the past 5 years, whether his ability to keep up with the pace of his job had changed, and whether the respondent felt more or less fatigued at the end of the day compared to 5 years earlier.

In the 1990 survey, several questions were addressed to Older Men sample persons evaluating their entire work life. This series asked about the respondent’s general satisfaction with his career, whether his work was mainly a source of income or was enjoyable in itself, and what aspect of his work the respondent found most enjoyable. Years and reference numbers for the



evaluation of work experience questions are found in the “other work attitudes” columns in the tables at the end of this section.

Attitude toward women working. Respondents were periodically asked a series of questions regarding their attitude toward the employment of women, particularly married women with children. Reference numbers for these questions are provided in the tables at the end of this section. In 1967, the Older Men responded to a single question that asked which of five statements best described their feelings about a married woman with children ages 6–12 taking a full-time job outside the home. The statements were as follows: she should never work, it’s okay only if absolutely necessary to make ends meet, it’s okay if family would like extra income, it’s okay if she prefers to work, or she should work.

In the 1971 survey of Older Men, these attitudes questions were changed into a three-statement series. On a five-point scale from “definitely all right” to “definitely not all right,” respondents were asked how they felt about a married woman with children ages 6–12 taking a full-time job outside the home if it was absolutely necessary to make ends meet, if she wanted to work and her husband agreed, and if she wanted to work but her husband did not like the idea.

This same series of three statements was read to Young Men respondents in 1971, 1976, and 1981. However, the Young Men questions referred to a married woman with preschool-aged children rather than older children. In 1981, a second set of questions was added to the Young Men survey about the employment of wives in general, without the presence of children specified. On a five-point scale from “strongly agree” to “strongly disagree,” respondents reported how they felt about eleven statements (e.g., a woman’s place is in the home, not in the office or shop; employment of both parents is necessary to keep up with the high cost of living; men should share the work around the house with women).

Other work-related attitudes. In every survey through 1976, Young Men respondents were asked what type of work they hoped to be doing at age 30. These occupations were coded using the Census Bureau coding system; in addition, some surveys include codes for the occupation using the various indices described in the “Occupations & Occupational Prestige Indices” section of this guide. Desired occupation at age 30 questions are not included in Table 4.17.2



below; these variables can be easily located by searching on “Age 30” in the Young Men data set.

Finally, two surveys asked Young Men respondents who had moved between interviews whether they were more or less satisfied with their employment now compared to their employment before the move.

User Notes: Cross-cohort analyses of job satisfaction and attitudes toward women working questions are possible using items from the Mature and Young Women cohorts and the NLSY79.

Related Variables: The “Health” section of this guide describes psychological well-being questions not directly related to employment. The “Retirement” section discusses questions about retirement attitudes asked of the Older Men. The “Job Characteristics Index” and “Discrimination” sections include information about specific work-related issues. The Knowledge of the World of Work scale, addressed to Young Men, is explained in the “Aptitude, Achievement & Intelligence Scores” section. Questions regarding reservation wages and the respondent’s reaction to hypothetical job offers are described in the “Wages” section.

Survey Instruments & Documentation: These questions are located within the “Current Labor Force Status,” “Work Experience and Attitudes,” “Attitudes toward Work,” and “Retrospective Work History” sections of the Older and Young Men questionnaires. More information on constructing the job satisfaction scale can be found in Appendix 23 of the Young Men Codebook Supplement.

ReferenceAndrisani, Paul J.; Appelbaum, Eileen; Koppel, Ross; and Miljus, Robert C. “Work

Attitudes and Labor Market Experience: Evidence from the National Longitudinal Surveys.” Philadelphia, PA: Center for Labor and Human Resource Studies, Temple University, 1977.



Table 4.17.1 Reference Numbers for Older Men Work Attitudes Questions

Survey Year

Satisfaction with current job

Factors liked best / least about job

Current job comp. to previous job

Attitude toward women working Other work attitudes

1966 R00131., R00550. R00132.–R00137. — —motivation R00138., R00169.commitment to work R00139.,

R00160., R00551.

1967 R00751., R00767. R00752.–R00757.R00702., R00703., R00768., R00769., R01047.–R01050.

R00775. —

1969 R01600., R01272., R01258. R01259.–R01264. R01273. — —

1971 R01988. R01989.–R01994. R01986., R01987. R02001.–R02003., R02522.

desired hrs R01702., R01703.progressed R02054.–R02060.pressure/pace R02077.–R02079.

1976 R02997. — — —desired hrs R02932., R02933.progressed R03086.pressure/pace R03099.–R03102.

1978 — — — — desired hrs R03778., R03779.1980 R04116. — — — commitment to work R04115.1981 R04594. — — — —1983 R05536. (curr./last) — — — —

1990 R06055., R07150. (curr./last) — — — evaluation of work life R06158.–

R06162.

Table 4.17.2 Reference Numbers for Young Men Work Attitudes Questions

Survey Year

Satisfaction with current job

Factors liked best / least about job

Current job comp. to previous job

Attitude toward women working Other work attitudes

1966 R00226. R00227.–R00232., R00602. — — motivation R00317.

1967 R00832., R00846. R00833.–R00838. R00847., R00848. — —1968 R01359., R01374. R01360.–R01365. R01375., R01376. — —

1969 R01943. R01944.–R01949. — — commitment to workR01957.–R01960.

1970 R02738. R02739.–R02744. R02751., R02752. — —

1971 R03507. R03508.–R03513. R03505., R03506. R03523.–R03525.

desired hrs R03207., R03208.commitment to work R03537.progressed R03580.–R03586.residence change R03685.

1976 — — — R04943.–R04945.desired hrs R04572., R04573.progressed R04991.pressure/pace R04995.–R04998.

1978 R05535., R05563. R05536.–R05541., R05564.–R05569. — — —

1980 R06027. R06028.–R06033. — — —

1981 R06910. (curr./last) R06911.–R06916. — R07724.–R07737.

commitment to workR07904.–R07907.

residence change R08010., R08035.


Chapter 4: Topical Guide—Labor Force Status

4.18 Labor Force Status

The following section describes the various labor force status variables present in the NLS. It does not provide either a comprehensive discussion of all questions asked in the “Current Labor Force Status” sections of the various NLS survey instruments or a thorough treatment of the detailed information available on labor market transitions and work histories. Users should consult the table of contents of this guide for references to additional labor market–related topics of interest, e.g., work experience, job characteristics, job satisfaction, industries, occupations, wages, etc.

Each questionnaire’s “Current Labor Force Status” section collects information on the labor market activity in which respondents were engaged during most of last week. This series replicates the questions asked in the monthly Current Population Survey (CPS) of American households conducted by the U.S. Census Bureau for the U.S. Department of Labor. The primary purpose of the CPS is to collect up-to-date information about the number of persons in the country who are employed, unemployed, or not looking for work during a given survey week. Results from the CPS surveys, released in the monthly publication Employment and Earnings, provide detailed information, classified by age, sex, race, and a variety of other characteristics, on the employment and unemployment experience of the U.S. population.

Older and Young MenA series of variables are available, for each cohort, on respondents’ labor force status during the survey week. In addition to these respondent-specific variables discussed below, data are available on the work experiences of other family and household members.

Survey Week Labor Force StatusThe following three sets of variables are available on each respondent’s labor force status during the survey week:

1. Activity Most of Survey Week: The ‘Activity Most of Survey Week’ variables reflect each respondent’s reply to the survey question “What were you doing most of last week?” “Last week” refers to the full calendar week (Sunday through Saturday) preceding the date of



interview. Although coding categories differ by cohort, the following categories of responses have been used to classify data within each cohort: (a) working, (b) with a job—not at work, (c) looking for work, (d) going to school, (e) unable to work, and (f) other. An additional coding category, “retired,” is present for all survey years of the Older Men; “keeping house” was added as a coding category for the 1990 Older Men survey.

Definitions for each of these labor market activities are intended to be consistent with those utilized in the CPS. Charts found later in this section provide definitions of key CPS labor force concepts; Census interviewers are instructed to use their CPS manual for assistance in coding the current labor force status questions. Due to the fact that Census is responsible for CPS data collection, it is likely that NLS CPS questions are interpreted in a consistent manner.

The main survey week activity question is followed by a second question that seeks to identify those respondents who did any work at all last week in addition to a main survey week non-working activity (such as “looking for work” or “going to school”); this follow-up question is asked of all respondents except those who indicate that they were working or were unable to work.

2. Employment Status Recode (ESR): ESR is a variable created by the Census Bureau that recodes responses to various employment-related questions into a consistent and more accurate measure of each respondent’s survey week labor force activity. A series of decision rules, depicted in Table 4.18.1 below, cluster information collected from ten questions dealing with, for example, main survey week activity, hours worked, whether/why absent from a job, job search activity, occupation, class of worker, etc. into positive or negative indicators of “working,” “with a job but not at work,” and “unemployed (looking for work).” In order to be assigned to one of these recodes, a respondent must display at least two positive and no negative indicators that he belongs to one of these groups; otherwise he is considered to belong to one of the “not in the labor force” categories. More detail on the decision pathways used



to assign each recode and on exceptions to the general rules presented below can be found in “Standardized Employment Status Questions and Recodes” (Census 1977).

ESR is available for all survey years except for the years in which telephone interviews were conducted. Information on creation inconsistencies can be found in the User Notes below as well as within each cohort’s codebook.

3. Labor Force Group Variables: Check items present in the questionnaires of the Older and Young Men provide a series of summary variables indicating the labor force group to which a respondent belonged, e.g., working, looking for work, unable to work, retired, or in the Armed Forces. Variables which compare current survey year’s labor force group with that of the previous survey year and which link labor force group to other variables such as school enrollment, presence of children under age 18, marital or retirement status are available for select survey years and cohorts.

Survey Instruments: Questions on main survey week activity are located at the beginning of the “Current Labor Force Status” sections of each questionnaire. The labor force group variables appear as check items throughout the questionnaires.

Documentation: Each year’s Interviewer’s Reference Manual provides detailed instructions for interviewers on how to code this section of the questionnaire in a manner consistent with CPS. Decision rules that guide Census in its creation of the ‘Employment Status Recode’ variables can be found in “Attachment 5: Standardized Employment Status Questions and Recodes (Census 1977)” in the Codebook Supplements.



Table 4.18.1 Employment Status Recode Creation: Older and Young Men

Ten Employment-Related Questions Used to Create ESR/MLRMajor activityWhether worked last weekHours worked

Whether absent from jobWhy absent from jobWhether looking for workWhat doing to find work

Reason could not take job (availability for work)OccupationClass of worker

ESR/MLR - 1 WORKING ESR/MLR - 2 WITH A JOB, NOT AT WORK

ESR/MLR - 3 UNEMPLOYED (LOOKING FOR WORK)

Positive indicators

1. Working last week2. 15+ hours worked3. Class of worker entry other

than “never worked”4. 1–14 hours worked combined

with class of worker entry other than “without pay”

1. Absent from job or business2. Class of worker entry other

than “without pay” or “never worked”

3. Reason for absence from work other than “layoff” or “new job to begin in 30 days”

1. Absent from job or business2. Reason for absence is “layoff”

or “new job to begin in 30 days”3. Looking for work4. Any entry in class of worker

(including “never worked” and “without pay”)

5. Method of looking for work entered other than “nothing”

Negative indicators

1. 1–14 hours worked combined with class of worker = “without pay”

1. Reason for absence from work is “layoff” or “new job to begin in 30 days”

2. Working last week3. Any hours worked4. Class of worker is “without

pay”

1. Method of looking for work is “nothing”

2. Not available for work3. Reason for absence from work

is “other” (not “layoff” or “new job to begin in 30 days”)

4. Working last week5. Any hours worked

Source: Census Bureau. “Standardized Employment Status Questions and Recodes.” Washington, DC: U.S. Department of Commerce, April 1977. This document is distributed by CHRR as “Attachment 5: Employment Status Recodes” and is an important source of information on exceptions to the general indicators listed above.

User Notes: The various codes that categorize activities for those respondents not in the labor force vary across survey years and cohort. Tables 4.18.2 and 4.18.3 below present the coded values by survey year for the Older and Young Men ESR variables.



Table 4.18.2 Employment Status Recode Codes: Older Men

1966 1967 1968 1969 1971 1976 1981 1990Working 1 1 1 1 1 1 1 1With a Job, Not at Work 2 2 2 2 2 2 2 2Unemployed 3 3 3 3 3 3 3 3In School 4 4 4 4 4 4 4 4Retired 5 5 5 5 5 5 5 5Unable to Work 6 6,71 6 6,71 6,71 6 6 6Blank 7 7Other 8 8 8 8 8 7 7 8Keeping House 7Never Worked 0 0 0 0 0Noninterview (Blank All Years)

1 Two codes were used in these surveys to distinguish between years. For example, the 1967 “Unable to Work” code was separated into “6–Unable to work in 1967 only” and “7–Unable to work in 1966 and 1967.”

Table 4.18.3 Employment Status Recode Codes: Young Men

1966 1967 1968 1969 1970 1971 1976 1981Working 1 1 1 1 1 1 1 1With a Job, Not at Work 2 2 2 2 2 2 2 2Unemployed 3 3 3 3 3 3 3 3In School 4 4 4 4 4 4 4 4Blank 5 5 5 5 5 5 5 5Unable to Work 6 6 6 6 6 6 6 6Blank 7 7 7 7 7 7 7 7Other 8 8 8 8 8 8 8 8Never Worked 0 0 0 0 0 0 0 0Noninterview Blank All Years

User Notes: While the “Current Labor Force Status” sections of NLS questionnaires follow the wording and format of those asked in the CPS, users should be aware that NLS CPS sections include additional questions over and above those found in the CPS surveys.

ESR has been traditionally used by many researchers to restrict the sample of those respondents interviewed in a given survey year. Users should be aware that this procedure cannot be used with data from the 1990 Older Men cohort and should consult the codebook notation for ESR.



Researchers examining employment over time can construct a loose approximation of ESR by using positive responses to the following three questions: (1) Did you do any work at all last week? (2) Did you have a job or business from which you were temporarily absent? and (3) Have you been looking for work? A respondent, for example, who is coded “Other” on ‘Activity Most of Survey Week’ but has a job from which s/he was absent would be reclassified as “working.”

Labor Force Status TablesThe tables in the following section are organized as follows: Tables 4.18.4 and 4.18.5 depict the labor force status, as measured by ‘Employment Status Recode’ variables, across survey years for the Older Men sample as a whole and for respondents by race. Tables 4.18.6 and 4.18.7 provide similar information for the Young Men. Readers should note that these tables contain unweighted frequencies. They should only be used as an aid in describing raw frequency counts in these data and must not be used to make inferences about population data.

Table 4.18.4 Labor Force Status: Older Men Civilian Respondents 1966–90

Survey Year 1966 1967 1968 1969 1971 1976 1981 1990Total Interviewed 5020 4744 4648 4381 4175 3487 2832 2092

In the Labor Force

Total 4673 4377 4221 3942 3550 2148 1070 298Employed 4595 4295 4131 3884 3469 2044 1032 291Unemployed 78 82 90 58 81 104 38 7

Not in the Labor Force

Total 347 367 427 439 625 1339 1762 1611In School 0 1 0 2 2 0 0 0Retired 42 28 89 62 186 915 1440 1469Unable to Work 219 276 314 302 327 331 178 107Other 81 60 24 72 109 93 144 31Keeping House — — — — — — — 4Never Worked 5 2 0 1 1 — — —

ESR not available — — — — — — — 183Not Interviewed 0 276 372 845 845 1533 2188 2928Year Total 5020 5020 5020 5020 5020 5020 5020 5020



Table 4.18.5 Labor Force Status: Older Men Civilian Respondents by Race 1966–90

Survey Year 1966 1967 1968 1969 1971 1976 1981 1990Non-Black Respondents

Total Interviewed 3600 3402 3338 3141 3016 2521 2077 1603

In the Labor Force





Black RespondentsTotal Interviewed 1420 1342 1310 1240 1159 966 755 489

In the Labor Force







Table 4.18.6 Labor Force Status: Young Men Civilian Respondents 1966–81

Survey Year 1966 1967 1968 1969 1970 1971 1976 1981Total Interviewed 5225 4790 4318 4033 3993 3987 3695 3398

In the Labor Force



Total 1605 1286 938 705 552 388 170 106In School 997 994 751 527 383 253 67 21Unable to Work 5 9 12 9 12 23 37 37Other 71 81 109 143 1371 112 66 48Never Worked 532 202 66 26 20 0 0 0

Not Interviewed 0 435 907 1192 1232 1238 1530 1827Year Total 5225 5225 5225 5225 5225 5225 5225 5225

1 Includes two respondents recoded from “Code 7” to “Code 8-Other.”

Table 4.18.7 Labor Force Status: Young Men Civilian Respondents by Race 1966–81

Survey Year 1966 1967 1968 1969 1970 1971 1976 1981Non-Black Respondents

Total Interviewed 3787 3477 3155 2976 2962 2998 2848 2622

In the Labor Force




Not Interviewed — 310 632 811 825 789 939 1165Year Total 3787 3787 3787 3787 3787 3787 3787 3787

Black RespondentsTotal Interviewed 1438 1313 1163 1057 1031 989 847 776

In the Labor Force




Not Interviewed — 125 275 381 407 449 591 662Year Total 1438 1438 1438 1438 1438 1438 1438 1438

1 Includes one respondent in each racial group recoded from “Code 7” to “Code 8-Other.”


Chapter 4: Topical Guide—Marriage & Fertility

4.19 Marriage & Fertility

Older MenMarital status. Questions on marital status have been asked of respondents in each survey year, except for the 1968 mail survey. In general, the resulting ‘Marital Status’ variable includes six coding categories: married—spouse present, married—spouse absent, widowed, divorced, separated, and never married. In some years, the raw variable contains only five categories (combining married—spouse present and married—spouse absent). Researchers are encouraged to use the revised marital status variables available for these years; these revised variables provide a six-category distribution comparable to other years.

A single question in 1990 assessed the sample person’s attitude about his marriage. This interview asked the respondent whether he thought of himself and his wife as “two separate people” or “a couple.”

Marital transitions. Limited marital transition information was collected in the later Older Men surveys. In 1976, the respondent reported the date of his first marriage and, if applicable, the date he was widowed or divorced. In 1981, the survey first asked if the respondent had been married more than once. He then provided the dates of his first and most recent marriage, as well as the date he was most recently divorced or widowed. The 1990 interview of sample persons was the same as the 1981 survey, except that the date of first marriage question was dropped. All of the marital history variables can be located in the data set by searching for the phrase “Marital Status & Background.”

In 1990, if the widow of a deceased sample person was interviewed, she was asked to report the marital status of the respondent at the time of his death (e.g., whether she and the sample person were married, divorced, separated, etc.), as well as the year she and the sample person were first married.

Characteristics of the respondent’s spouse. Information on the respondent’s spouse was collected in the household roster during all interviews except the 1968 mail survey; these data are described in detail in the “Household Composition” section of this guide. In several of the early surveys, created variables are available that summarize the information on respondents’ wives



collected in the household record so that researchers do not need to sort through the roster data. In addition, the 1981 survey included a special set of questions about the work experience of the respondent’s wife. Data collected include whether the wife had ever worked; the start and stop years, industry, and occupation of her longest job; whether she had worked in the previous 12 months; her plans for retirement; and whether she expected to receive retirement income from Social Security or an employer pension.

A large amount of information is available for widows of Older Men respondents interviewed in 1990. This questionnaire collected information about the widow’s work experience, current marital status, household composition, income, assets, and residence in a long-term care facility.

User Notes: No specific questions on partners were asked of the Older Men; the household roster is a possible source of limited partner information. Although the list of possible relationships to the respondent on the “Household Roster” section of the questionnaire did not include “partner” in the early years, the relationship codes were revised in later years to include this category.

Children. If the respondent’s children resided in his household, data were collected in the household roster section of all questionnaires except 1968 regarding their age, education, and employment. This information is described in detail in the “Household Composition” section of this guide. In 1966–75, survey staff used the information from the household record to create variables providing the total number of the respondent’s children in the household; in some years, additional created variables break this total down into groups based on age, education, and employment status. A second series of questions in the 1971 interview asked about the respondent’s educational expectations for the youngest son and daughter living in his household and currently enrolled in school.

Every survey except 1968 asked the respondent how many dependents he had, excluding his wife. In personal interviews, the respondent stated his relationship to dependents outside his household. An additional series of questions in the



1981 survey asked the respondent if anyone was likely to become dependent on him in the future and, if so, their relationship to him.

More limited information was gathered about children who did not live with the respondent. The initial survey asked for the total number of such children, whether each was a son or daughter, and the highest grade completed by each child. A detailed data collection in 1976 recorded information about all children outside the household in a roster format. The respondent provided each child’s age and relationship to the respondent. If the child was age 16 or older, the survey collected his or her highest grade completed, labor force status in the past year, and occupation. The respondent also reported the marital status of each daughter, as well as her husband’s occupation.

The 1981 and 1990 interviews collected information about the frequency of contact between the respondent and children living outside his household. In 1981, the survey first recorded the total number of children living elsewhere; the 1990 survey gathered the total number of living children, regardless of residence. Both of these interviews asked a series of questions about how often the respondent visited, called, or corresponded with his children. In 1990, the respondent also stated how long it would take the closest child to get to the respondent’s residence.

Related Variables: Data were regularly gathered regarding the health of the respondent’s wife; these questions are discussed in the “Health” section of this guide. The “Pensions, Social Security & Retirement” section contains additional information about the respondent’s contact with his children, as well as a description of questions asking to whom he would turn for help with a financial or emotional problem.

Young MenMarital status. Questions on marital status have been asked of respondents in every survey. In most years, the resulting variable uses the same six coding categories as the Older Men. The Young Men marital status variables for 1973 and 1975 contain only five categories, not differentiating married—spouse absent from married—spouse present. Additionally, a ‘Marital Status and Family Status’ variable, created during several early survey years, combined information about marital status and the presence of children.



Marital transitions. Marital history data were collected from the Young Men during the final four surveys of that cohort. The 1976 survey, the first to include marital history questions, recorded the total number of marriages the respondent had, the date (month and year) of his first marriage, how and when the first marriage ended, the date of his current or most recent marriage, and the date he was divorced, separated, or widowed after his most recent marriage. In 1978 and 1980, married respondents reported the date of their current marriage and respondents who were divorced, widowed, or separated provided the date that event occurred. The final survey of the Young Men recorded the dates and types of up to four changes in the respondent’s marital status since the 1980 interview date (or the 1978 interview date if the respondent was not interviewed in 1980).

Characteristics of the respondent’s spouse/partner. As with the Older Men, most information on the wives of Young Men was collected in the household roster. These data are described in the “Household Composition” section of this guide. Additionally, the 1981 interview contained a separate series of questions referring to the respondent’s wife or partner. Data collected include birth date; whether the wife or partner had been enrolled in school since the 1980 interview date (or the 1978 interview date if the respondent was not interviewed in 1980); for enrolled wives and partners, the highest grade completed; weeks worked in the past 12 months; hours per week usually worked; occupation; and weeks spent looking for work. Researchers should note that the wife was not included on the household roster in 1981, so this series of questions replaces her roster data.



User Notes: Users are urged to examine the original questionnaires to determine exact wording, context, universe, and coding categories. In addition to the differences already mentioned, there are many related variables such as marital status collected retrospectively for noninterview years and interviewer check items that use different categories than those described above. When marital transitions were updated from a midpoint of a previous year rather than from a previous interview, certain vital information may be missing. For instance, if a respondent from the Young Men cohort was interviewed in 1976, not in 1978, but was interviewed again in 1980, his marital history was updated since a specified date in 1978 (not 1976). If he was married in 1976 but divorced and remarried before 1980, his marital status would be married for both 1976 and 1980 and he would have no marital transitions recorded. His wife, however, would be a different person with different characteristics than in 1976. It is imperative to examine the questionnaires to determine exactly what information is recorded.

For surveys before 1981, the “Household Roster” is a possible additional source of partner information. Although the list of possible relationships to the respondent on the “Household Roster” section of the questionnaire did not include “partner” in the early years, the relationship codes were revised in later years to include this category.

Children. If the respondent’s children resided in his household, age, education, and employment data about each were collected in the household roster section of each questionnaire. This information is discussed in the “Household Composition” section of this guide. In 1966, survey staff used the information from the household record to create variables providing the total number of the respondent’s children in the household and the number in various age groups.

Very limited information was gathered about children who did not live with the respondent. The 1981 survey asked for the total number of children the respondent had ever had and the birth date of each child not residing in the household. Additionally, every survey asked the respondent how many dependents he had, excluding his wife. In all surveys except 1973 and 1975, the respondent stated his relationship to dependents outside his household; these answers can be used to obtain some information about the number of children



supported by the respondent. The questions about dependents were not asked of respondents still living with their parents during the 1966–68 interviews.

The 1981 Young Men survey contained several questions related to fertility expectations and attitudes. The questionnaire first asked for the number of children respondents thought would be ideal and the number they wanted. After reporting how many children they had already, respondents stated how many more children they expected to have and when they expected to have their next child.

Household chores. The 1981 interview included a detailed series of questions regarding the respondent’s participation in various household activities. The respondent was presented with a list of common chores (e.g., cooking, laundry, grocery shopping, household paperwork, child care) and asked whether he had none, some, about half, most, or all of the responsibility for each chore. He then stated the total number of hours per week that he spent performing household tasks.

User Notes: The series of questions on household chores was asked of both Mature and Young Women in multiple surveys. Researchers can use these data to compare perceptions about household responsibilities across genders and generations.

Related Variables: Information on the health of the respondent’s wife was gathered in several surveys; these data are described in the “Health” section of this guide.

Survey Instruments & Documentation: Current marital status of the respondent was generally transcribed from the updated Household Record Cards to page one of the questionnaire or to the Information Sheet. In some survey years, however, current marital status is collected in other sections of the questionnaire, such as “Health” or “Work Attitudes.” The derivations of the revised versions of marital status for the Older Men are listed in the codebook. Marriage and fertility information for respondents in the Older Men cohort was collected in the “Marital History” or “Family Relationships” sections. Marital transitions and child data for the Young Men were collected in the “Marital



History,” “Family Members,” “Family Background,” or “Marital Status” questionnaire sections.


Chapter 4: Topical Guide—Military

4.20 Military

Older MenActive Service: The members of this cohort were asked in the 1967 and 1976 surveys whether they had ever served in the U.S. Armed Forces. If so, the dates of service were collected; these were specified by time periods in 1967 (e.g., World War II, Korean War, peacetime after January 1955) and by month and year in 1976.

Training: The original Older Men survey, conducted in 1966, asked the respondents about any vocational training programs they attended while in the Armed Forces. Follow-up questions determined whether the respondent completed the program, the type of training, the length of the program, and whether the respondent used this training at his current or last job.

Related Variables: The military was also included in Older Men surveys as a source of disability payments, pension or retirement income, and medical care payments. For more details about disability and pension income, see the “Pensions, Social Security & Retirement” section of this guide. Medical care payments the military made to the respondent or his wife, including payments for care at a long-term facility, are discussed in the “Health” section.

Survey Instruments & Documentation: These questions are found in the 1966 “Education and Training,” 1967 “Family Background,” and 1976 “Marital History and Other Background” questionnaire sections.

Young MenBecause of the younger age of this cohort and the Vietnam War, considerable data were collected about respondents’ military service. Armed Forces questions were included in all Young Men surveys except the 1973 and 1975 telephone surveys.

Active Service: In the 1966, 1969, 1971, 1976, and 1981 surveys, respondents who had ever served in the Armed Forces reported the branch they served in. These surveys asked how respondents entered the Armed Forces (e.g., drafted, enlisted as a regular) and how many months they spent on active duty. Respondents were also asked for the military occupation they had held for the



longest time and whether they were an officer or an enlisted man at that time. The final Young Men survey in 1981 asked whether respondents were ever in combat during their service, as well as the duration of that combat.

Separation from active duty was also a focus of the military questions. The 1966, 1969, and 1971 questionnaires collected the respondent’s age at the time of separation. The respondent’s rank and the date the separation occurred were recorded in 1969, 1971, 1976, and 1981.

For those who had never served in the U.S. Armed Forces, the 1966 survey questioned whether they had ever tried to enter active military service. If so, a follow-up question asked for the reason they were not accepted (e.g., failed physical exam, failed written test). Similarly, the 1967–71 surveys recorded the current draft classification of each age-eligible respondent and, if applicable, the reason for their 1-Y or 4-F rejection.

Training: Details about training received in the military (other than basic training) were gathered in the 1966, 1969, 1971, 1976, and 1981 surveys. Respondents attending such programs reported whether they completed the program (in 1966, 1969, and 1971 only); the length of the program; and whether they used this training on their current or last job. In addition, the general training questions of the 1976, 1978, 1980, and 1981 surveys included the Armed Forces as a possible provider of a training course.

Benefits/Disability/Income: The later Young Men surveys included questions concerning income received from disability or Veteran’s benefits. The 1976 questionnaire asked whether the G.I. Bill was an important element in the respondent buying his first home; in 1978 and 1980, the G.I. Bill was included in a question about other income. More detailed military income questions were asked in 1976 and 1981. These included the type of G.I. or Veteran’s benefits the respondent had ever used (e.g., housing, medical care, vocational training); if he had a disability or medical discharge; whether he had a disability rated as service-connected by the Veterans Administration; and the compensation rating for that disability. Additionally, the 1981 survey questioned whether the respondent or his wife/partner received income in the past 12 months from Veteran’s compensation or pension as a result of a disability or illness.



Related Variables: Several questions about education and work decisions for this cohort included participation in the military in the list of possible responses. Education questions included why respondents ended their education at a particular point (1966), the reason for changing their college attendance plans (1968–71), why they were not currently attending college (1968–71), and why they decided to get more training (1968 and 1969). Various surveys included participation in the Armed Forces or waiting to be called into military service as reasons for not looking for work or for not working. The first Young Men survey in 1966 asked the respondents what they would do if they were to permanently lose their present job tomorrow; one possible response was “enter the Armed Forces.” The military was also connected to a respondent’s work history by a question in 1971, 1976, and 1981 that asked whether the respondent’s period of military service helped or hurt his career and the reason why.

Survey Instruments & Documentation: Questions about the military in the Young Men’s surveys can be found in the “Educational Status,” “Current Labor Force Status,” “Work Experience and Attitudes,” “Family Background,” and “Assets and Income” questionnaire sections.


Chapter 4: Topical Guide—Occupations & Occupational Prestige Indices

4.21 Occupations & Occupational Prestige Indices

Older and Young MenData on the occupation that respondents were seeking or in which they were employed or received training were collected during most survey years of the Older and Young Men. Some surveys gathered information on the occupation of intervening and dual jobs.

In each survey, an open-ended question asked respondents “What kind of work (are/were) you doing?”. Follow-up questions fielded during some survey years elicited more specific information on job duties and job title. Verbatim responses from the respondent were entered onto the questionnaire by the interviewer and then coded by Census personnel using the Census Bureau alphabetical index of occupations and industries. All occupation variables for the Older and Young Men regardless of survey year are coded with three-digit 1960 Census codes. Additionally, the 1983 and 1990 interviews of Older Men doublecoded occupation of current or last job using both the 1960 and three-digit 1980 classifications.

A series of edited variables are available for respondents in the Older and Young Men cohorts that provide a three-digit and one-digit occupational code for the current or last job ever reported by the respondent. These variables can be differentiated from the direct pick-up ‘Occupation of Current or Last Job’ variables by the absence of a question number in the source field. The User Notes in the “Industries” section of this guide provide more information on the editing and creation procedures used for certain occupation variables.

User Notes: Variable titles for occupations listed within the various NLS documentation items do not always specify the Census coding system utilized. In such cases, users should assume that the 1960 classification was applied.



In the questionnaires and Census versions of the data files provided to CHRR, the responses to some employment-related questions were coded in such a way as to require reference to the response to another question. For example, in the Older Men survey, the occupation of longest job was coded “same as current job” if the respondent’s job at the 1966 survey date was the longest job he had held between leaving school and the 1966 survey. Otherwise, the actual three-digit occupation code was coded. In the CHRR data files, no such cross-referencing is required. In the above illustration, “occupation of longest job” is given the three-digit code of “current job” in those instances where current and longest job are the same. Relevant notations are present within each cohort’s codebook.

Within the Older and Young Men cohorts, “job” changes are tracked with ambiguity as to whether this would be an occupation change, employer change, or both.

Duncan Index. All three-digit 1960 Census occupational categories have been assigned a two-digit ordinal prestige score based upon the education and income distributions of the occupation. The scores, ranging from 0 to 97, may be interpreted either as estimates of prestige ratings or simply as values on a scale of occupational socioeconomic status (Duncan 1961). A series of created variables provide the Duncan Index score for the current or last job ever reported by the respondent. The universe for these variables is all respondents interviewed in a given survey year for whom occupational data have been collected.

GED and SVP scores. The 1966–71 surveys of the Young Men include created variables providing two special occupational scores for current or desired occupations: a General Education Development (GED) score and a Specific Vocational Preparation (SVP) score (Department of Labor 1965, Appendix B). The GED score is a representation of the amount of general education or life experience needed to perform a given job. The score represents three factors: reasoning development, mathematical development, and language development; each of these factors is divided into six levels, with one representing the least amount of education and six the most. The first number in the 3-digit GED score represents the level of reasoning required for the job, the second number is the



level of mathematical achievement, and the last number indicates language requirements.

The SVP score considers the opposite proposition: that some amount of time is required to learn to perform a specific occupation at an average level of competence. This single-digit score ranges from 1 to 9, with 1 meaning that the job only requires a short demonstration, 2 indicating that the job requires up to 30 days of training and experience, and so on up to 9, which means that the job requires more than 10 years of specific learning and experience before it can be performed at an acceptable level.

Job family and job level. The 1966–71 Young Men interviews also provided classifications for the respondents’ current or desired occupations according to job family and job level (Scoville, 1969). Using the first three digits of the Census Bureau occupational codes, the job family variables assign each job to one of 18 occupational clusters (e.g, machines and equipment—specialized, inspection, farming, clerical, research and design, etc.). Based on the GED and SVP scores and the average earnings for each occupation in 1960, the job level variables provide information about the comparative levels of skills, training, and abilities required for different jobs. Each job is classified in levels I-V, with class ‘I’ jobs requiring the greatest amount of skill and knowledge and class ‘V’ jobs the least.

Survey Instruments & Documentation: Questions on occupations are found within the “Current Labor Force Status,” “Work History,” and “Retirement and Pension” sections of the questionnaires; occupation of household members has been collected as part of the “Family Background” or “Household Members” sections. Attachment 2 in each cohort’s Codebook Supplement provides the 1960 & 1980 (Older Men only) Census of Population industry and occupational classification codes and the accompanying Duncan Index scores. Appendix 19 in the Young Men Codebook Supplement lists the GED, SVP, job level, and job family codes for each 1960 occupation code.

Background information on the development of the 1960 and 1980 classification systems and the relationships between the 1960 and 1970 coding categories and between the 1970 and 1980 codes is available within various Census publications (Census 1972, 1989).



ReferencesCensus Bureau. 1960 Census of Population Alphabetical Index of Occupations

and Industries (Revised Edition). Washington, DC: U.S. Government Printing Office, 1960.

Census Bureau. “1970 Occupation and Industry Classification Systems in Terms of Their 1960 Occupation and Industry Elements.” Technical Paper 26. Washington, DC: U.S. Government Printing Office, 1972.

Census Bureau. 1980 Census of Population Classified Index of Industries and Occupations. Washington, DC: U.S. Government Printing Office, 1980.

Census Bureau. “The Relationship Between the 1970 and 1980 Industry and Occupation Classification Systems.” Technical Paper 59. Washington, DC: U.S. Government Printing Office, 1989.

Census Bureau. U.S. Census of Population: 1960. Subject Reports. Occupational Characteristics. Final Report PC (2)-7A. Washington, DC: U.S. Government Printing Office, 1960.

Duncan, O.D. “A Socioeconomic Index for All Occupations.” In: Occupations and Social Status, A.J. Reiss, Jr. et al. New York: Free Press, 1961.

Scoville, James G. The Job Content of the U.S. Economy 1940–1970. New York: McGraw-Hill, 1969.

U.S. Department of Labor. Dictionary of Occupational Titles, Third Edition. Volume II: Occupational Classification and Industry Index. Washington, DC: U.S. Government Printing Office, 1965.


Chapter 4: Topical Guide—Pensions, Social Security & Retirement

4.22 Pensions, Social Security & Retirement

Older MenDecisions about retirement were a main focus of the Older Men surveys. This section first discusses questions about the actual retirement and disability income received by Older Men respondents and their family members. It then describes the series of questions addressed to respondents still in the work force about their expected income and pension benefits after retirement. Finally, this section considers the wide variety of questions asked of respondents regarding their retirement plans, experiences, and attitudes.

Actual pension, retirement, and disability incomeSome information about pension income in the previous calendar year was collected in every survey except 1968. In general, most surveys asked about income from a government pension and income from other sources, such as a private employer pension. However, there were variations in the amount of detail recorded and the family members in the household to whom the questions applied. Similarly, every survey except 1968 asked about receipt of Social Security or Railroad Retirement benefits, although the family members referred to in the questions varied across surveys. Tables 4.22.1 and 4.22.2 summarize the topics and universes of the Older Men pension and Social Security income data collection.

Table 4.22.1 Older Men Pension Income Data

Survey year Universe Data collected1966 Any family member Total amount from federal, state, and local government pensions

1967, 1969 Any family member Total amount from government pensionsTotal amount from all other pensions

1971, 1976Any family member; amounts recorded separately for respondent, wife, and other members

Total amount from government pensionsTotal amount from all other pensions

1973, 1975 Respondent only Total amount from all pensions

1978 Respondent and wife; amounts recorded individually

Total amount from all pensions; sources recorded but amounts by source not specified

1980 Respondent and wife; amounts recorded individually

For each of following sources, amount received and whether it had increased, decreased, or remained the same in past 2 years:

local government private employerstate government unionArmed Forces personal planother federal government other source

1981 Respondent and wife Other family members

Same as 1980Total amount from all sources

1983, 1990 Respondent and wife (or widow and Same as 1980



current husband in 1990)

In addition to the information described in the table, the 1966 survey asked whether any family member was eligible for a pension from a source other than the government; actual income amounts from such a pension were not collected. For the respondent only, the 1971 and 1976 other retirement pension series included a series of follow-up questions that identified how many employer-provided pensions the respondent was receiving. For the pension providing the largest income, this series then gathered the age that pension receipt began, the number of service years completed when benefits began, whether retirement occurred voluntarily, and whether the pension would have been larger had work continued with that employer. Specific information on the type of pension plan (e.g., defined benefit, defined contribution) was not collected for this cohort.

Table 4.22.2 Older Men Social Security/Railroad Retirement Incomeand Disability Income Data

Survey year Universe Social Security/Railroad Retirement data Disability income data

1966, 1967 Any family member Total amount Amount in each category, recorded for respondent and for all other family members combined:

Social Security DisabilityVeteran’s compensationWorkers’ compensationAid to the Blind/Permanently DisabledOther

1969, 1971 Any family member Amounts recorded separately for respondent, wife, and other members Same as 1966/1967

1973, 1975 Respondent only Total amount —

1976, 1981 Any family member Amounts recorded separately for respondent, wife, and other members

Same as 1966/1967, but Aid to the Blind/ Permanently Disabled dropped

1978, 1980 Respondent & wife Total combined amount —

1983 Respondent & wife Total amount each received individually Same as 1976/1981

1990Respondent & wife (or widow & current husband)

Total combined amount if received combined check; otherwise recorded separate amounts

Same categories as 1976/1981; amount of each type received recorded separately for respondent and spouse

A set of retrospective questions was fielded in 1983 that gathered information on whether the respondent and/or his wife had ever applied for or ever received Social Security Disability benefits, whether benefits had ever been or were currently being received, and the year such benefits were last received. The 1990 survey of living respondents and of widows of deceased respondents also recorded



information regarding on whose work record Social Security/Railroad Retirement benefits were based.

User Notes: Researchers should note that imputed income values are available for 476 of the 503 respondents who were missing Social Security/Railroad Retirement benefit amounts from the 1981 data collection; see variables R05271.10 and R05477.10.

For all income categories, including pensions, Social Security, and disability payments, the 1990 sample person and widow questionnaires gathered information on total amounts received during both the 1989 calendar year and the month before the interview.

Expected pension and retirement incomeSeveral surveys collected information about the income respondents expected to receive after retirement. The initial survey asked whether respondents would be eligible to receive Social Security benefits and whether they would receive retirement benefits from another source. If so, the type of source was recorded (e.g., government, private employee, personal plan, military).

Expected pension benefits for those respondents who were working for a private employer or the government were explored in more detail in 1969, 1971, and 1976. The 1969 survey simply asked whether the respondent expected to receive income from a pension provided by his employer. An expanded series in 1971 and 1976 elicited anticipated retirement ages when the respondent would be eligible for full and reduced benefits, as well as the corresponding monthly income amounts. The respondent was also asked if he would be eligible to receive a benefit and the amount if he were to leave his job today or to retire today. In addition to expected pension benefits, the 1971 and 1976 surveys collected information about the respondent’s expected income from Social Security/Railroad Retirement. The respondent reported whether these benefits would be a source of income for him (and his wife) when he reached retirement age; if so, an expected monthly income amount was gathered. Finally, the 1971 and 1976 surveys asked respondents to report the total amount of monthly or yearly retirement income they expected to receive from all sources.



The 1981 survey collected the most detailed information about expected retirement income. In this interview, respondents who had not yet retired were read a list of 10 possible sources of income and asked whether they expected to receive income from each source. Possible income sources included Social Security/Railroad Retirement, current employer pension, past employer pension, personal pension plan, investments, money from children or relatives, Supplemental Security Income, other welfare payments, rental income, or another source. The respondent then reported whether he was eligible to receive Social Security benefits and the amount per month he and his wife expected to receive. Also fielded in 1981 was a series of questions on the respondent’s wife’s retirement plans and whether she expected to receive income from Social Security (based on her own work record) or from an employer-provided pension; separate amounts for the wife were not collected. Finally, considering all of these sources, including Social Security information already provided, respondents reported their total expected retirement income on either a yearly or monthly basis. Follow-up questions requested information on how much of this total income would come from a current employer pension, a past employer pension, or from Social Security/Railroad Retirement benefits.

User Notes: Researchers should be aware that the same list of 10 sources of retirement income was addressed to two different universes of respondents based on whether they planned to retire; the two series of questions must be combined if information about all respondents is desired.

Retirement attitudes and experiencesPlans for retirement. Older Men respondents have regularly answered questions about when they intended to retire and what factors affected that decision. In every survey except 1968 and 1990, respondents reported the age at which they expected to retire. Five surveys (1967, 1969, 1971, 1980, and 1983) asked respondents whose plans had changed since the previous interview about the reason for the difference. Finally, every personal interview except 1990 asked employed respondents whether their employer had a compulsory retirement age, whether they would work longer than that age if they could, and whether they expected to retire earlier than the mandatory age.



Limited information was gathered about the retirement plans of the respondent’s wife. In 1981 and 1983, married respondents stated their wife’s expected retirement age. The 1981 survey further asked whether the respondent and his wife had made decisions jointly or separately about when to retire.

Respondents in the Older Men cohort answered questions in two surveys about training courses taken in preparation for retirement. In 1971, respondents who participated in any kind of training stated whether they expected to use the training after they retired. The 1981 interview asked a series of questions specifically tailored to retirement training courses, including whether the respondent had ever taken such a course, the duration of the course, and whether he had found it to be helpful.

Finally, respondents were asked several questions about financial obligations that might affect retirement plans. In 1971 and 1976, the respondent stated the year when he expected to have no dependents other than his wife. The 1971 survey also asked whether the respondent wanted to leave an inheritance to his children.

Timing of actual retirement. In later surveys, retrospective information was collected about the retirement decisions made by respondents. The final five surveys included a series of questions asking why the respondent retired, whether the respondent’s employer encouraged him to retire, whether the respondent would change the age at which he retired in retrospect, and whether he would have continued to work if his employer had permitted it. In 1980–90, respondents reported whether they had ever retired and the year this occurred, even if they had subsequently reentered the work force.

Attitudes toward retirement. Older Men respondents were asked a number of questions about their attitudes toward retirement. The reference numbers for these questions and the years in which each was asked are provided in Table 4.22.3.



Table 4.22.3 Reference Numbers for Older Men Retirement Attitudes Questions

Question topics (R = respondent) 1971 1976 1978 1980 1981 1983 1990Is R looking forward to retirement R02110.

Agree/disagree with statements about retirement (e.g., retirement is a pleasant time of life, work is the most meaningful part of life)

R03861.–R03865.

R04250.–R04254.

R05024.–R05028.

Wife’s attitude about R’s retirement R02111.

Friends’ attitudes about their own retirement R02116.

Is life in retirement better/worse than R expected R03177. R03860. R04228. R04803. R05690. R06155.

Things liked best/least about retirement R04844.–R04850.

In addition, several questions in the 1971 survey were combined to create an index (R02528.) of the respondent’s overall attitude toward retirement. Included in this index are the three 1971 questions in the above table, a question about the ages at which the respondent’s retired friends left the workforce, and the question asking whether the respondent wanted to leave an inheritance to his children.

Social networks and activities. Respondents regularly provided information about their plans for retirement. The 1966 survey asked whether they intended to take another job after retiring from their current employer and the number of hours per week they would like to work. Seven subsequent surveys contained a more general question about plans after retirement; respondents were asked to select all that applied from a list including travel, relax, enjoy a hobby, volunteer, or work with or without pay. If the respondent planned to work, follow-up questions asked about the type of work he would like and the number of hours per week he would work.

User Notes: Like other multiple response (“check all that apply”) questions, the retirement plans questions were coded using a geometric progression format. Program statements to unpack such variables are presented in Appendix C.

Three interviews asked respondents about their leisure time activities. In 1978, these questions were addressed to all respondents; in 1981, they were asked only of respondents who were retired; and in 1990, these data were gathered from all respondents not living in an institution. Although the exact list of



activities varied among surveys, it typically included sports or exercise, reading, working at hobbies, visiting friends or relatives, and volunteer work. Follow-up questions asked for the amount of time spent at each activity.

Respondents were asked about their social and familial relationships in 1981 and 1990. A series of questions in both of these interviews asked about sources of emotional support for the respondent other than his spouse. In addition to the total number of people with whom the respondent had a close emotional relationship, the surveys recorded the relationship of the closest confidante, the distance to his or her residence, and the frequency of contact between the respondent and this person. A similar series of questions asked to whom the respondent would turn if he had a serious financial problem. In 1990, an additional pair of questions gathered information about the person who most often helped the respondent with daily tasks like grocery shopping or housecleaning. Finally, both the 1981 and 1990 interviews asked the respondent to describe how often he visited or talked with children who did not live in his household.

Survey Instruments & Documentation: Retirement and pension questions were asked in the “Assets and Income,” 1966 “Retirement Plans,” 1971 and 1976 “Plans for the Future,” and 1981 “Retrospective Work History” questionnaire sections; the “Income” section of the 1990 sample persons questionnaire; and the “Income” and “Medical Care Prior to Death” sections of the 1990 widow’s questionnaire.

Young MenBecause they were still fairly young when interviews with this cohort were discontinued, no information on eligibility for or income from pensions and Social Security was collected for these respondents. Limited information on disability income was gathered and is discussed below.

Social Security Disability/Other Disability Payments. During most survey years, disability income was included within the “income from other sources” questions—question wording did not differentiate disability income from rental, interest, or dividend income. In 1966–71 and 1976, this question about other income referred to both the respondent and his wife and asked about “income as a result of disability or illness.” The 1973 survey, which referred only to the



respondent’s income, asked the respondent to report the total amount of other income and up to three sources; coding categories included “disability pension including Workman’s Compensation” and “Social Security insurance, gifts from family members, royalties, and inheritances.” The 1975, 1978, and 1980 surveys did not include any specific references to disability income but did contain an “other income” question. Finally, the most detailed information was collected in 1981. This survey asked for the amount of income received from Veteran’s compensation or pension, Workers’ Compensation, Social Security disability, and any other disability payments. Amounts were recorded separately for each type of payment and for the respondent and his wife.

User Notes: Like other multiple response (“check all that apply”) questions, the 1973 other income sources question was coded using a geometric progression format. Program statements to unpack such variables are presented in Appendix C.

Survey Instruments: The “Assets and Income” section of the relevant questionnaires.

Related Variables: The “Geographic Mobility” section of the 1981 Young Men questionnaire collected information on the effect of the respondent’s move to his current residence on (1) the seniority rights of the respondent or spouse and (2) the retirement plans of the respondent or spouse. Coding categories delineated whether the respondent/spouse had lost some, none, or all seniority or pension/retirement rights or whether he or she had no seniority or retirement rights before the move. The fringe benefit series included “retirement pension program” as one of the benefits made available by a current or past employer to respondents of the Young Men cohorts. Availability should not be confused with actual coverage under a pension plan or receipt of pension benefits.


Chapter 4: Topical Guide—Poverty Status & Public Assistance Support Sources

4.23 Poverty Status & Public Assistance Support Sources

Older and Young MenData on public assistance income sources have been collected for the Older and Young Men during select survey years. In general, data are available on income received from public assistance/welfare, Aid to Families with Dependent Children (AFDC), food stamps, or Supplemental Security Income. Users should be aware that there is considerable variation across years in the types of public assistance income sources for which data were collected. In addition, (1) universes (all family members, any family member, respondent and spouse, respondent only, spouse only), (2) reporting periods (past calendar year, previous 12 months, most recent month), and (3) the wording of questions differs substantially both within and across cohorts.

Data on housing assistance were collected for select survey years of the Older Men. The following series of poverty status and income-poverty level ratio variables are available for the Older Men cohort only: (1) Poverty status variables created for the 1966, 1967, and 1969 surveys indicate

whether the income of a respondent’s family unit was above or below a set of Census-developed poverty income guidelines based on number of children, farm-nonfarm residence, and marital status of the respondent.

(2) A ratio of the respondent’s family income in the previous year to the poverty level was created for the 1966, 1967, 1969, 1971, 1973, and 1975 surveys.

Survey Instruments & Documentation: Questions on public assistance income sources are found in the “Assets and Income” or “Income” sections of the Older and Young Men questionnaires. Appendix 1 in the Older Men Codebook Supplement provides the poverty guidelines used in creation of the poverty status variables. Variable creation procedures for the Older Men ratio variables can be found within the codebook.

User Notes: NLS surveys also collect data on Unemployment Insurance, Workers’ Compensation, Disability and Social Security. None of these sources of income are considered here as part of “public assistance.” The “Pensions, Social Security & Retirement” section of this guide describes some of these additional income sources.


Chapter 4: Topical Guide—Poverty Status & Public Assistance Support Sources

Table 4.23.1 Public Assistance Questions by Survey Year, Type of Assistance,Reference Period & Universe: Older and Young Men

Type of AssistanceSurvey Year

Welfare / Public Asst. Food Stamps SSI / Public

Assistance Public Housing Reference Period Universe

Older Men1966 R00277. R00279. — — Past Calendar Year Family Members

1967 R00825. R00828. — — Past Calendar Year Family Members



1973 R02674. — — — Previous 12 Months Respondent

1975 R02832. — — — Previous 12 Months Respondent

1976 R03340. R03343. R03336.–R03338. — Past Calendar Year Family Members

1978 R04030. R04025. R04024. — Past Calendar Year Respondent/Wife

1980 R04341. R04337. R04283. R04342.1 Past Calendar Year Respondent/Wife

1981 R05281. R05276. R05273.–R05275. R05279. Past Calendar Year Family Members

1983 R05949. R05945. R05936. R05950. Previous 12 Months Respondent/Wife

1990 R06776., R07547. R06767., R07538. R06762., R07533. R06774., R07545. Past Calendar Year Respondent, Spouse

Young Men1966 R00374. — — — Previous 12 Months Family Members

1967 R00872. — — — Previous 12 Months Family Members





1975 R04321. R04318 — — Previous 12 Months Family Members

1976 R05331. R05332. R05313./R05326.2 — Previous 12 Months Family Members

1978 R05896.3 R05890. — — Previous 12 Months Respondent/Wife

1980 R06373.3 R06368. — — Previous 12 Months Respondent/Wife

1981 —4 R07888. R07896. — Previous 12 Months R or Wife/Partner

1 The universe for this question is family members.2 The first question refers to the respondent’s receipt; the second refers to his wife’s receipt.3 These refer to the monthly average amount for the previous 12 months.4 The 1981 Young Men survey included question R07893., which asked about the respondent’s or wife/partner’s monthly average of AFDC

receipt over the previous 12 months. This is the only question referring to AFDC receipt for either the Older Men or the Young Men.


Chapter 4: Topical Guide—Race, Ethnicity & Nationality

4.24 Race, Ethnicity & Nationality

Older and Young MenOne race variable is available for each respondent in the Older and Young Men cohorts. Nationality of the respondent was created from data collected on parents’ and grandparents’ birthplaces.

Race/Ethnicity: In the Older and Young Men cohorts, ‘Race’ (R00023. for both cohorts) is a three-category variable (black, white, and other) available for the respondent only and, in general, is derived from the household screening. According to the Current Population Survey Interviewer’s Reference Manual (Census 1962) in use at the time of the screening, race was to be determined by interviewer observation. Interviewers were instructed to code Mexicans, Puerto Ricans, and other Latin Americans as “white” unless they were obviously of another race and were to include Japanese, Chinese, American Indian, Korean, Hindu, Eskimo, etc. in the “other” category. Table 4.24.1 presents a distribution of race by nationality for the Older and Young Men. At the time of the first survey of each cohort, race information for each respondent was manually transferred to the questionnaire from information that had been entered on the Household Record Cards during the 1966 household screening. (Only in the case of the creation of a new household, where a respondent had moved out of the household in which he was living at the time of the screening, would the interviewer fill out a new Household Record Card, in which case all household member information would be newly recorded.)

Table 4.24.1 Race by Nationality: Older and Young Men

Nationality1Older Men Young Men

Total Non-black Black Total Non-black BlackUS or Canada 2877 1588 1289 3939 2527 1412

North or West Europe 1049 1044 5 86 80 6

Central or East Europe 364 363 1 523 520 3

South Europe 250 248 2 307 307 0

Latin America 39 37 2 252 251 1

Other 117 100 17 106 94 12

NA 324 220 104 12 8 4

Total 5020 3600 1420 5225 3787 1438



1 Countries included in each category are listed in Table 4.24.2. The U.S. and Canada category appears over-represented because nationality was based on birthplace of parents and grandparents, so it includes all whose parents and grandparents were born in the U.S. or Canada.

Nationality: The variable ‘Nationality of R,’ created during the initial survey year, is available for each respondent (R00584. for the Older Men and R00625. for the Young Men). The nationality of the Young Men respondents was derived from the first parent or grandparent born outside of the U.S. and Canada using the following decision rules: if the father was born outside of the U.S. and Canada, his nationality was assigned to the respondent; if he was born inside the U.S. and Canada but the respondent’s mother was not, her nationality was assigned to the respondent; and so forth. The derivation of the Older Men nationality is suspected to be similar, but it is unclear at this time. Categories for both cohorts include U.S. or Canada, North or West Europe, Central or East Europe, South Europe, Latin America, and other. There are no separate categories for Asian or African countries. Specific countries included in each category are not listed in the codebook with the nationality variable but are included in Table 4.24.2.

Table 4.24.2 Country Codes for the Nationality Variables

Coding Category Countries

North or West EuropeAustria, Belgium, Denmark, England, France, Germany, Iceland, Ireland (Eire), Luxembourg, Netherlands, Northern Ireland, Norway, Scotland, Sweden, Switzerland, Wales, Yugoslavia

Central or East Europe Albania, Bulgaria, Czechoslovakia, Estonia, Finland, Hungary, Latvia, Lithuania, Poland, Romania, USSR

South Europe Andorra, Azores, Gibraltar, Gozo, Greece, Italy, Liechtenstein, Malta, Monaco, Portugal, San Marino, Spain, Trieste, Vatican City

Latin America Mexico, Central America, South America

Related Topics: A single 1966 variable identifies whether the birthplace of Young Men respondents was in the United States (R00380.). ‘Which Foreign Language Spoken Regularly in Household When R Was Age 15’ (R03691.) is available for the Young Men in 1971.

Survey Instruments & Documentation: Race was recorded on Household Record Card form LGT-1, which was used at the time of the 1966 screening and the initial interview, and was manually transferred to the first page of the initial



year’s questionnaire. Birthplace was collected in the “Family Background” section of each cohort’s initial year’s questionnaire.

The Young Men codebook contains information on the specific derivation of the nationality variable.

User Notes: Although race of all household members was collected in the initial screening, these data never became part of the data set.

ReferenceCensus Bureau. “Current Population Survey and Housing Vacancy Survey:

Interviewer’s Reference Manual.” Washington, DC: U.S. Department of Commerce, 1962.


Chapter 4: Topical Guide—Training

4.25 Training

Older MenTraining questions were fielded during all personal interviews except the 1990 resurvey (i.e., 1966, 1967, 1969, 1971, 1976, and 1981). Information was collected during the initial survey on the training experiences of these then middle-aged men while in high school and since their regular schooling ended. Subsequent surveys updated that information. Questions on the training plans of respondents and their participation in pre-retirement planning programs were fielded during select survey years.

The 1966 survey first collected information on whether the respondent had been enrolled in any vocational or commercial training in high school. Respondents then reported whether, since completing regular schooling, they had obtained additional training from a business college/technical institute, a full-time (six weeks or more) company training school, the Armed Forces, or another vocational/technical/apprenticeship program. The survey also asked about general training in English, math, or science. For each training program, data were collected on the type of training (professional/technical, managerial, clerical, skilled manual, or general courses), the length of the training, and whether the respondent completed the program or used the acquired skills on his current job. For the longest vocational training taken outside of regular school, the Census Bureau created three variables that summarized the type of program, the sponsor or training provider, and whether that training was used on the respondent’s current job. Coding categories for the created training type and sponsor variables are consistent with those of the raw variables, with “apprenticeship” added as a separate type of training provider.

Subsequent surveys updated the respondent’s record from the date of the last personal interview. Data on additional training courses or educational programs in which the respondent had enrolled either on the job or elsewhere were gathered. In addition to the core information on each program (i.e., type, length, sponsor, completion status, and use on current job), information was collected on hours per week spent in the training, the reason the respondent did not complete, and the reason for enrollment in additional training. Created variables are present for the 1967, 1969, and 1971 surveys that summarize the total number of hours and



number of weeks between the current and last interview dates that the respondent spent in training. In 1971, a new set of questions was fielded on the respondent’s plans to take additional training courses in the near future. The 1976 survey added a question on whether the respondent expected to use this training in retirement; the 1981 interviews asked whether a pre-retirement planning program had been taken and, if so, who sponsored the program, the program’s duration, whether it had been completed, or whether the respondent found it helpful.

Table 4.25.1 below presents by survey year and race the number of Older Men respondents participating in training programs.

Related Variables: Two variables, created in 1966 and 1971, are titled ‘Skill Content Current Last Job’ and provide information on the usual length of training time necessary to perform the respondent’s current/last job.

Survey Instruments: The “Education & Training” sections of the 1966, 1967, 1969, and 1971 questionnaires and the “Retrospective Work History” sections of the 1976 and 1981 instrument.

Table 4.25.1 Numbers of Respondents Participating in Training Programsby Survey Year and Race: Older and Young Men 1966–81

YearOlder Men Young Men

Total Non-Black Black Total Non-Black Black1966 1846 1520 326 577 464 1131967 382 316 66 363 268 951968 1 1 1 564 448 1161969 434 367 67 651 523 1281970 — — — 767 613 1541971 413 342 71 776 613 1631973 1 1 1 1061 870 1911975 1 1 1 1095 912 1831976 437 377 60 944 781 1631978 1 1 1 1090 914 1761980 1 1 1 1061 910 1511981 252 216 36 8112 705 106

Note: The 1966 variable for both cohorts was constructed from four separate questions that asked the respondent whether he had ever participated outside of regular school in specific types of training programs. Thus the reference period for 1966 was “ever participated” while that of subsequent survey years was participation “since the last interview,” “in the past five years,” etc. Excluded from this table is participation in military training reported in separate question series by the Older Men in 1966 and by the Young Men during the 1966, 1969, 1971, 1976, and 1981 surveys.1 No training questions asked of the cohort in these years.



2 Does not include respondents who were enrolled in a training program at the 1980 interview date and reported attending that program between the 1980 and 1981 interviews. Information about these at-last-interview training programs was collected in a separate series of questions.

Young MenDuring the initial survey, respondents not enrolled in school who were not college graduates were asked a series of questions on their plans for more education or training. These respondents also reported whether they had ever been enrolled outside of regular school in business/technical training, a company training school, other vocational training (including apprenticeship) and whether they had taken additional general courses (e.g., English, mathematics, or science) since they stopped attending full-time school. For each training experience, information was collected on the type of training (technical/professional, managerial, clerical, skilled or semi-skilled manual, or other training including basic or general courses); number of months and hours per week spent in training; whether the program was completed and if not, the reason; and use of the skills acquired in the training program on the respondent’s current/last job. Each subsequent survey continued the core data collection with information on whether additional formal training or educational courses had been taken either on the job or elsewhere, the type of training, duration and intensity, completion status, reason for the training, and use on present job. These questions referred to the period since the date of the respondent’s last interview or since October 1 of the previous interview year if he had not participated in the previous interview. Changes in the eligible universe of respondents, new sets of questions, and changes in the coding categories of the core data collection are discussed below. Table 4.25.1 above presents the number of Young Men respondents participating in training programs.

Over time, the universe of respondents asked the training questions was expanded. In 1966 and 1967, only respondents who were not enrolled in school and who were not college graduates were asked about their training experiences. In 1968–76, all respondents not currently enrolled in regular school answered questions about training experiences, regardless of college graduation status. In 1978–81, all respondents, both enrolled and not enrolled, reported participation in training programs other than regular school.

Information on the type of institution providing the training was added beginning with the 1967 survey. Core coding categories for the type of school variables



include business college/technical institute, company training school, correspondence school, regular school, and other (including federally funded MDTA or Title V programs). Codes for apprenticeship and Armed Forces were added as permanent categories in 1973. During the 1971 and 1976 survey years only, the regular school category was differentiated into community or junior college, high school, and area vocational school. Information was collected during the 1976 survey on training programs offered by a community agency (church, YMCA, etc.).

The 1970 questionnaire included a retrospective on any training courses or educational programs taken prior to October 1967. During the 1971 and 1976 surveys, a series of questions on plans for future training courses or educational programs were added. The number of questions varied but included the type of training, reason, month and year such training was expected to be taken, and factors influencing the decision to seek training. The 1973, 1975, and 1976 surveys added a question to the core series that provided information on whether the respondent was still enrolled in the last survey year’s training program. Stop dates (month completed training) were collected during the 1967–71 and 1976 surveys.

Limited information was collected regarding apprenticeships and journeyman’s status. The 1975 interviews included a retrospective on whether the respondent had ever been enrolled prior to October 1971 in an apprenticeship program, start and stop dates, whether the program had been completed, and the type of trade learned. Journeyman’s status questions included in the 1978 survey collected information on whether the respondent held a journeyman’s card, the occupation of up to three journeyman’s trades, whether journeyman’s status was attained following an apprenticeship, and the year the first card was received.

User Notes: As noted above, the type of training provider question included an Armed Forces answer category. In addition to any information collected in the regular training series, respondents in the Young Men cohort answered a second series of questions referring specifically to military training. These questions are discussed in the “Military” section of this guide.

Created Variables: Summary variables were created from data collected in 1966, 1968, and 1971 that indicate whether the respondent completed or used



various types of training, e.g., ‘Completed or Used Professional or Technical Occupational Training as of 66?,’ ‘Completed or Used Clerical Occupational Training, 66-71?,’ etc.

Related Variables: The 1966 “CPS” section included a question on whether the respondent’s current job required more or less or about the same skills as the job held one year ago. The 1966 “Attitudes Towards Work” section included three hypothetical questions dealing with the type of training, the type of school, and the method of financing additional schooling should the respondent permanently lose his present job and decide to return to school or get training. The “CPS” section of the 1980 questionnaire included a set of questions on the training methods used to learn the respondent’s current/last job and, for those with more than one method, the most helpful method was identified. Coding categories included college courses, vocational school, company training, Armed Forces, apprenticeship, on-the-job training, promotion, relative/friend, informal training, etc. This 1980 series replicated a set of retrospective questions from a supplement to the April 1963 Current Population Survey that collected information on the training methods used to learn a specific job. Hills (1982c) presents comparison data from the 1980 NLS of Young Men and a sample of young people from the 1963 CPS.

Survey Instruments: The “Education & Training” section of the 1966 questionnaire; the “Educational Status” sections of the 1967–71, 1976, and 1981 questionnaires; and untitled sections of the 1973, 1975, 1978, and 1980 questionnaires.

Documentation: “Attachment 2: 1960 and 1980 Census Industrial & Occupational Classification Codes” in the Young Men Codebook Supplement contains the codes for the 1975 and 1978 apprenticeship training and journeyman trade variables.

User Notes: Rumberger (1984) states that participation in training may be understated due to a questionnaire skip pattern that omitted from questioning respondents enrolled in school during the fall fielding period who might have participated in training prior to their current school enrollment.



ReferencesFreeman, Richard B. “Occupational Training in Proprietary Schools and Technical

Institutes.” Review of Economics and Statistics 56,3 (August 1974): 310–18.

Grasso, John T. “The Contributions of Vocational Education, Training, and Work Experience to the Early Career Achievements of Young Men.” Ph.D. Dissertation, The Ohio State University, 1975.

Grasso, John T. and Myers, Steven C. “The Labor Market Effects of Investment in Human Capital.” In: Career Thresholds, Volume 6: A Longitudinal Study of the Educational and Labor Market Experience of Young Men. A.I. Kohen, et al. Columbus, OH: CHRR, The Ohio State University, 1977.

Hills, Stephen M. “How Craftsmen Learn Their Skills: A Longitudinal Analysis.” In: Job Training for Youth, R. Taylor, et al. Columbus, OH: National Center for Research in Vocational Education, 1982.

Hills, Stephen M. “A Longitudinal Analysis of the Monetary Benefits to Training for Craftsmen.” Columbus, OH: CHRR, The Ohio State University, 1982.

Hills, Stephen M. “Longitudinal Analyses of Training Processes in the United States.” Presented: Urbana, IL: Rupert P. Evans Symposium on Vocational Education, 1982c.

Hills, Stephen M. and Santos, Richard. “Perceived Utility of Job Training Methods among Young Men.” Columbus, OH: CHRR, The Ohio State University, 1983.

Lillard, Lee and Tan, Hong W. “Private Sector Training: Who Gets It and What Are Its Effects?” Research in Labor Economics 13 (1992): 1–62.

Mangum, Stephen and van Adams, Arvil. “The Labor Market Impacts of Post-School Occupational Training for Young Men.” Growth & Change 18,4 (Fall 1987): 57–73.

Mangum, Stephen and Ball, David. “Skill Transfer and Military Occupational Training.” In: The Changing Labor Market: A Longitudinal Study of Young Men. Stephen M. Hills, ed. Columbus, OH: CHRR, The Ohio State University, 1986.

Rumberger, Russell W. “The Incidence and Wage Effects of Occupational Training among Young Men.” Social Science Quarterly 65, 3 (September 1984): 775–88.

Rumberger, Russell W. “The Intensity of Occupational Training and its Effect on Earnings.” In: Market Defenses: Early Work Decisions of Today’s Middle-Aged Men. Stephen M. Hills, ed. Columbus, OH: CHRR, The Ohio State University, 1983.


Chapter 4: Topical Guide—Wages

4.26 Wages

This section overviews the rate of pay information that has been collected for one or more jobs held by the respondent since the last interview, e.g., the current or last job, a second or dual job, or various intervening jobs. Related variables, not discussed here, include whether and under what conditions extra pay was received, how overtime work was compensated, whether wages were set by a collective bargaining agreement, the hours or shift usually worked, and the respondent’s preference for working different hours for different pay. Data are also available for most survey years of each cohort on reservation wages (the minimum wage required to accept a job by those not in the labor force) and on rates of pay associated with hypothetical job offers.

Older MenRate of pay. All surveys of this cohort except those conducted in 1968, 1973, and 1975 collected information on the amount of money that respondents employed in the private or governmental sectors usually earned at their current or last job. The open-ended question, “How much do you usually earn at this job before deductions?” typically elicited an hourly dollar and cents amount or a dollar amount only plus a time unit, e.g., week, month, year, etc., to which the wage applied. From this information coupled with that on usual hours worked per week, CHRR calculated a set of variables entitled ‘Hourly Rate of Pay at Current or Last Job *KEY*.’ Excluded from the universe of these variables are those respondents reporting time units of day, piece work, or other, as well as those who were self-employed or working without pay in a family business or farm. During the 1971, 1976, 1981, and 1990 surveys, rate of pay information was also collected about jobs other than the current/last job—for example, the job held for the longest time in the past 12 months or the job held longest in the past five years.

Hypothetical job offers. In every survey except 1968, 1973, and 1975, respondents were asked about their reactions to hypothetical job offers. Most frequently, these questions were set in the context of a job offer in the same geographical area in which the respondent currently lived, but occasionally the questions referred to a different geographical area. The hypothetical job offer series was asked separately for different groups of respondents depending on their current labor force status (e.g., employed, unemployed, or out of the labor



force). Component questions in this series asked about the rate of pay required to accept the job offer, hours per week the respondent would be willing to work on the job (beginning in 1971), and the occupation required to accept the position. Users should consult the questionnaires for minor variations in text wording or universes. Researchers can easily find these questions on the CD-ROM by searching under the word “Hypothetical.”

Young MenRate of pay. Information on the usual earnings of respondents working in the private and government sectors was collected during each survey year. A series of ‘Hourly Rate of Pay for Current or Last Job *KEY*’ variables has been created for each survey year except 1973 and 1975. Universe restrictions are similar to those described for the Older Men. Rate of pay information is also available for intervening jobs reported during the 1970, 1971, 1976, and 1981 surveys and for a second (or dual) job in which a respondent was employed at the time of the 1971, 1976, and 1981 interviews.

Hypothetical job offers. Young Men were administered the same series of hypothetical job offer questions as Older Men. These questions were included in all surveys except 1973 and 1975.

Survey Instruments & Documentation: Rate of pay information was collected in the “Current Labor Force Status,” “Current Labor Force Status and Work History,” “Work Experience and Attitudes,” “Employment,” “Work Attitudes,” or “Retrospective Work History” sections of each cohort’s questionnaires. Derivations for most created hourly rate of pay variables are presented within the cohort-specific codebooks.

User Notes: Derivations for select hourly rate of pay variables contained statements that set to “NA” values above and below designated extreme values. This truncation was not consistently applied either across cohorts or across the survey years of a given cohort. For example, school enrollment status—in addition to extreme pay values—was factored into the creation of the 1966 Young Men *KEY* variable. Derivations for certain created rate of pay variables do not appear within the public codebook or Codebook Supplement; users needing this information should contact NLS User Services.



ReferenceBureau of Labor Statistics. Work & Family: Changes in Wages and Benefits

Among Young Adults. Report No. 849. Washington, DC: U.S. Department of Labor, July 1993.

Parsons, Donald O. “Summary Tables: The National Longitudinal Surveys of Older Males 1966–1983.” Columbus, OH: CHRR, The Ohio State University, 1987.

Parsons, Donald O. “Summary Tables: The National Longitudinal Surveys of Young Males 1966–1981.” Columbus, OH: CHRR, The Ohio State University, 1987.


Chapter 4: Topical Guide—Work Experience

4.27 Work Experience

Although the NLS has collected information on labor force behavior since its inception, only partial work histories for respondents in the Older and Young Men cohorts can be constructed for certain survey years. The degree of completeness of the work history data varies by cohort and survey year.

For those wishing to measure labor force attachment over time, three approaches are available. One can examine: (1) the amount of time in weeks that a respondent spent working, unemployed (looking for work), or out of the labor force; (2) the start and stop dates of each job a respondent has held (a continuous job history); or (3) the start and stop dates associated with each employer for whom a respondent worked (a continuous employer history).

In general, summary weeks data (information on the number of weeks working, unemployed, and out of the labor force) were collected during each interview for either the previous 12 months or the previous calendar year. The term “summary weeks data” refers to the respondent’s answers (in weeks) to the following types of questions: “During the past 12 months, in how many different weeks did you do any work at all?” Respondents who worked 52 weeks were asked, “Did you lose any full weeks of work during the past 12 months because you were on layoff from a job or lost a job?” Respondents who worked less than 52 weeks were asked, “In any of the remaining weeks, were you looking for work or on layoff from a job?” Those answering “yes” were asked to report the number of weeks. Respondents who did not work during the past 12 months were asked if they had spent any time looking for work or on layoff and if they had, how many weeks. While placement and wording of the individual questions have varied, this core set of summary questions is always present in each interview.

Unfortunately, such data collection consistency did not occur in obtaining information to track all job and/or all employer changes. The gaps in information collected on weeks worked (see discussion below) are minor compared to the gaps in information on jobs held and employment spells. Due to the fact that personal and telephone interviews used different time reference periods, it is only possible to construct a complete job and/or employer record for the early survey years of the Young Men. During the early survey years, the Young Men were personally



interviewed every year because they were experiencing a great amount of change (movement between school and work, between jobs for an employer, or between employers). As long as this interview pattern was maintained, data to construct a continuous job history are available. With respect to employer change, it is almost always possible to link employers within a given interview; however, it is not possible to link all employers across interview years.

There are three different ways to construct a summary measure for number of weeks worked, seeking work, or out of the labor force. Users can examine the start and stop dates associated with each job, especially in the personal interview years, when the questionnaire included a detailed work history in a column format. (These variables can be found on the CD-ROM under the keyword “Most_Recent Job.”) When the information about start and stop dates is combined, a fairly complete picture of total number of weeks in the labor force can be pieced together. This is the procedure that has been used at CHRR to create the *KEY* weeks variables. Users attempting to create number of weeks worked themselves instead of using the created *KEY* variables need to pay close attention to the skip patterns followed in the early survey years. Many check items sent respondents to different parts of the questionnaire to answer questions worded specifically for their particular situations. When constructing number of weeks worked, users should pay particular attention to the dates in the detailed work history section. During the early survey years, the Census Bureau truncated the date the respondent started the job to the preceding interview date if it started before then, so the actual start date may not be available; in the later years, when an interviewer inadvertently gathered information before the date of the last interview, this information was sometimes left on the data file instead of being blanked out and eliminated.

Two alternatives to this time-consuming procedure of piecing the record together from start and stop dates include: (1) use of information from the summary weeks questions present in the questionnaire for all years or (2) a combination of data from the *KEY* summary weeks variables for those years in which they were constructed and information from the summary weeks questions for those years in which no *KEY* variable is available. The *KEY* variables (those variable with titles of ‘# of Weeks_Worked [reference period] *KEY*,’ ‘# of Weeks_Unemployed [reference period] *KEY*,’ and ‘# of Weeks_OLF [reference period] *KEY*’) were



created for all survey years in which respondents were personally interviewed, except for the 1990 Older Men resurvey. Care should be taken to check that the number of cases on the summary weeks variables is reasonably close to the number of respondents interviewed (since all respondents should have a value on these variables). If this is not the case, the user needs to make sure that the desired information is not present in another part of the questionnaire or to adjust for the fact that in some years respondents who had not worked since the last interview are assigned to “NA” or missing instead of being assigned a “zero” for zero weeks of works, as one would expect.

Gaps in the reference periods for the summary week variables occur in the early 1970s when the project phased in an alternating personal and telephone interview pattern. The regularly fielded personal interviews conducted during the early survey years gave way to a 2-2-1 interview pattern (i.e., two telephone interviews occurring two years apart followed by a personal interview at the end of the five-year period). The intent of the telephone interview was to obtain a brief update of information on each respondent and to maintain sufficient contact such that the lengthier personal interview could be completed. Due to the fact that the reference period for the summary weeks questions within a telephone interview was the previous 12 months and that no interview was conducted the year before each telephone survey, gaps in the summary weeks records of Older and Young Men respondents occurred. However, questions were added to the Older Men personal interview instruments that gathered information on weeks not worked over a five-year period, including the two missing years. By using simple subtraction, the total number of weeks worked over the five year period can be calculated. This additional information is not available for the Young Men.

The discussion below reviews by cohort the types of summary weeks information that are available from the questionnaire. Included is information on changes in the reference periods for which these data were collected. The weeks worked accounting is not completely accurate due to the slight over- or under-counting of weeks that occurs when a respondent is not interviewed exactly one year from the date of the last interview. If the respondent accurately answers the question on the number of weeks worked in the last (or past) 12 months and it had been 13 months since the last interview, the summary weeks variables would miss four weeks of employment status information. Although Census was asked in the early



years to interview each respondent as close as possible to the date of the previous interview, the actual dates of interview can and should be checked.

Older MenA nearly complete work history for weeks worked and a complete listing of the most important or longest held jobs are available for respondents in the Older Men cohort.

During the initial 1966 survey, each respondent was asked detailed questions regarding the following four jobs: (1) his current job or the last job he had held since 1961, (2) the job he had held before his current job or before his last job if it started after 1961, (3) the longest job of all the jobs he had ever had, and (4) the first job he had held for at least a month after stopping school full-time. By looking at the start and stop dates for these jobs, it is possible to obtain a global measure of weeks worked up to 1961 and a measure of weeks worked between 1961 and the date of the 1966 interview. Summary weeks questions, e.g., the number of weeks working, weeks unemployed, and weeks out of the labor force, were asked of the respondent for the calendar year 1965.

The 1967 questionnaire included questions on the current or last job for those respondents who had worked at all since June 15, 1966. Information was also gathered on one intervening job. If the respondent had held more than one such job, information was gathered on the job with the longest tenure. Respondents were also asked the summary weeks questions; however, the reference period on these questions was the last 12 months, not the previous calendar year as it was in 1966.

The 1968 mail survey elicited information on the respondent’s current job or his last job since June 1, 1967. The questionnaire also asked the respondent to report the total number of intervening jobs in the past 12 months and to describe the characteristics of the longest intervening job. Summary weeks questions refer to the last 12-month period.

In the 1969 survey, respondents were asked about their current job or their last job since June 1, 1968. If the respondent had changed jobs since the last interview, he was again asked for the total number of intervening jobs and for details about the



job held the for the longest time. The summary weeks questions asked about the last 12-month period.

The 1971 interview first collected information about the respondent’s current job and a dual job, if applicable. This survey then used a column format to collect information about all jobs held since the date of the last interview. The summary weeks questions also referred to the date of the previous interview. This means that, except for those respondents who were not interviewed in all years, a reasonably accurate number of weeks worked between 1965 and 1971 can be obtained.

Gaps in the employment record start with the 1973 telephone interview. In this shorter survey, respondents were asked questions only about their current job or the last job held since August 1971. The summary weeks questions were asked about the previous 12 months. The 1975 telephone interview followed a similar pattern, collecting information about the respondent’s current job or his last job since August 1973. The summary weeks questions again refer to the previous 12-month period.

In the 1976 personal interview, any respondent who had worked since August 1971 was asked for information on his current or last job, the job he held before his current or last job, and the job he had held for the longest time since 1971. There are two different sets of summary weeks questions. The usual set of summary week questions was asked for the previous 12-month period; another set collected information on weeks not working for the previous five years. These variables, ‘# of Weeks Not Working, 71-76,’ ‘# of Weeks Unemployed, 71-76,’ and ‘# of Weeks OLF, 71-76,’ can be used to patch the gaps left by the telephone interviews.

The 1978 telephone interview asked the respondent about his current job or the last job he had held since the 1976 interview date (or since August 15, 1976, if the respondent was not interviewed in 1976). The summary weeks questions asked about the last 12 months. The 1980 telephone interview repeated the 1978 pattern.

The 1981 personal interview was similar to the 1976 survey. Respondents provided information about their current or last job, the job they had held for the longest time in the past 12 months, and the job they had held for the longest time



since August 1976. As in the 1976 survey, there were two sets of summary weeks questions, one series asking about the previous 12 months and the second series referring to the period since August 1976.

The 1983 telephone interview mainly repeated the 1978 pattern; respondents provided information about their current job and the longest job held since the 1981 interview (or since August 15, 1981, if not interviewed in 1981). Respondents then answered summary weeks questions referring to the previous 12 months. However, the first set of summary weeks questions was followed by a second series that asked about the 12 months before that, so that the 2-year period between the 1981 and 1983 surveys was fully covered (depending on exact interview dates).

The 1990 reinterview included two questionnaires. The questionnaire addressed to living Older Men respondents (or their proxies) asked all respondents about their current or last job. Respondents who were institutionalized or who had not worked at all since their last interview were skipped past the remaining work experience questions. Those who remained in the eligible universe provided information about the job held the longest since their last interview, if it was different from the current or last job, and about the number of weeks worked each year from 1983 to 1989. Respondents who had not worked 52 weeks in 1989 were asked if they had been looking for work or on layoff during weeks not working.

The 1990 widow questionnaire collected more limited information about the sample person’s work experiences. The widow (or her proxy) first reported the date the respondent stopped working at the job he held at his previous interview date. She then answered questions about the longest job held by the respondent between his previous interview and his death and about the last job the respondent held before his death.

For those respondents still alive in 1990 (and surveyed in every personal interview year), summary weeks information on weeks working, weeks unemployed, and weeks out of the labor force is available for the years 1966 to 1983 and for 1989. Information on weeks worked is also available for the period 1983 to 1988. For deceased respondents who consistently participated in the survey, summary weeks information is available for 1966 through 1983. Employment start and stop dates collected about deceased respondents during in the widow interview can be



used to provide some information about the number of weeks worked, although sample persons who held more than three jobs between 1983 and their death are missing information about some jobs.

Young MenThe 1966 survey first asked if respondents were employed in the week before the survey. If not, respondents were asked to report when they last worked for two consecutive weeks, with responses grouped into the following categories: “never worked at all,” “never worked 2 weeks or more,” “before 1961,” or the month and year if the date was 1961 or later. Respondents then described their current job or the last job held since January 1, 1961. If the respondent was not enrolled in school, he also provided details on the job he held during his last full year in high school (if applicable) and the first job he held for a least one month after he stopped attending school full-time. Finally, all respondents who had ever worked were asked the summary weeks questions on weeks worked, unemployed, and out of the labor force for the previous 12 months.

In 1967, respondents first provided information about their current job or the most recent job held since October 15, 1966, were asked about that job. If applicable, respondents also provided information about their longest intervening job. All respondents except those unable to work were asked the summary weeks questions for the last 12-month period (respondents who had not worked were asked job search and layoff questions to distinguish between weeks unemployed and weeks out of the labor force). The 1968 and 1969 personal interviews followed the 1967 pattern.

The 1970 interview first collected data on the current or most recent job and then gathered a detailed work history regarding all jobs held since October 1, 1969. The summary weeks questions covered the period since October 1, 1969. In 1970, respondents who were not interviewed in 1969 were administered a special questionnaire that asked about the previous two years. In these cases, the same information was gathered as in the regular questionnaire, but the reference date was October 1, 1968.

The 1971 interview collected data on the respondent’s current job and a dual job, if applicable. Using a column format, the survey then gathered a detailed work history of all jobs since the date of the most recent interview. The summary weeks



questions similarly refer to the date of the previous interview. A retrospective series of questions, addressed only to respondents who had not been enrolled in school or enlisted in the Armed Forces at any survey point (slightly more than half the respondents), asked for the total number of weeks not working since October 1966. A follow-up question elicited information on how many of the weeks not working the respondent was on layoff or looking for a job.

The gaps in the employment record start with the 1973 telephone interview. In this survey, respondents described either their current job or the job most recently held since October 1971. The summary weeks questions referred to the last 12 month period. The 1975 telephone interview followed the 1973 pattern.

In the 1976 personal interview, which was similar to 1971, respondents provided information about their current job or the last job held since the date of the 1975 survey (or since October 1, 1975, if the respondent was not interviewed in 1975). Respondents were then asked about all other jobs held since 1975 and about the longest job held since November 1971. The summary weeks questions, addressed to all respondents, referred to either the 1975 interview date (or to October 1, 1975, for noninterviews in 1975) or the date when they stopped working at their current or last job.

The 1978 telephone survey asks the respondent to describe his current job or the last job held since the 1976 interview date (or since October 1, 1976, if the respondent was not interviewed in 1976). The main set of summary weeks questions referred to last 12 months as in previous surveys. A new question then obtained more limited information on the number of weeks worked for the 12-month period previous to the last 12 months. Respondents reported whether they had worked most of the year (46–52 weeks), more than half a year (26–45 weeks), less than half a year (1–25 weeks), or had not worked at all. By using the midpoints of the ranges, researchers can approximate the number of weeks worked, although they cannot distinguish between weeks unemployed and out of the labor force. The 1980 interview repeated the 1978 pattern.

In the final 1981 interview, data were collected for those who had worked since the date of the 1980 interview, if they were interviewed in 1980, or since October 1, 1980, for those not interviewed in that year. Respondents were asked for information on both their current or last job and all other jobs held since 1980.



Respondents were also asked about the longest job held since November 1976. The summary week questions were asked of all respondents; the pattern replicated that used in 1976.

The detailed work history for the Young Men focuses on jobs rather than employer changes. Summary week variables are available for all but four years; range data on weeks worked are available for two of these four years.

ReferencesParnes, Herbert S., et al. The NLS Older Male Sample Revisited: A Unique Data

Base for Gerontological Research. Columbus, OH: CHRR, The Ohio State University, 1992.

Parnes, Herbert S. and Sommers, David G. “Shunning Retirement: Work Experience of Men in Their Seventies and Early Eighties.” Journal of Gerontology 49, 3 (1994): S117–S124.

Descriptive TablesThe tables below are presented to provide the researcher with information on sample sizes by race and interview year for weeks worked and number of employers. Due to the restructuring of the 1990 interview, tables for the Older Men cover the years 1966–83. The labor force attachment of Older Men interviewed in 1990 is described in Parnes et al. (1992, 1994).

Four tables are provided for each of the men’s cohorts. Tables labeled “Number of Weeks Worked All Survey Years” provide information on the average number of weeks worked by respondents interviewed in all survey years. These two tables summarize all information available on number of weeks worked. Tables labeled “Number of Weeks Worked/Year by # of Survey Years Reported Work” present information on the average number of weeks worked by number of years that the respondent was interviewed. Tables labeled “Number of Weeks Worked by Survey Year” depict the number of weeks worked for each respondent interviewed in that survey year, including both the mean and number of cases for those who report work and those who do not report any weeks worked. Tables labeled “Number of Employers by Survey Year” present information on the numbers of employers reported during each survey year.

A number of decisions were made during the construction of these tables in an attempt to make the information comparable across cohorts. The tables are not



weighted and are not intended to be used to make inferences about populations. The universe for the first two sets of tables is all respondents who were interviewed in all years. Years in which the *KEY* or summary week variables were found to have an upper range greater than 52 were truncated back to 52. In those years that a *KEY* variable covers a two year period, the total number of weeks was divided by two. The weeks tables do not take into account whether or not the respondent was in the labor force; if a respondent was interviewed and did not report any weeks worked, he was assigned a “zero” even if, for example, he was permanently handicapped and would not have been in the labor force under normal conditions. The number of respondents in the “NOT WORKING” categories in the third and fourth set of tables are similar although not identical. There was no attempt to eliminate respondents who did not have information available for both weeks and employers. The last set of tables presents information on the number of employers reported each survey year; however, the reference period varies between and across cohorts. “Survey year” could refer to the last twelve months, or to a period since the last interview that was one, two, three, or more years ago. Examining information on the total number of employers across time is difficult and time-consuming. Although it is possible to find information for most detailed work history years on the same and different employers within the survey period, the main linkage across years is the one for the current employer in the “CPS” section. In other words, it is not possible in the early survey years to know, for example, that the intervening employer in the second column of the detailed work history section is the same employer as that entered two years later in the third column of the work history without making a number of assumptions based on matching of the job and/or employer characteristics. In later survey years, it is possible to link an employer in a work history column to the employer at the time of the last interview. However, use of this extra information was beyond the scope of these tabular presentations.

Table 4.27.1 Number of Weeks Worked All Survey Years (Unweighted): Older Men

Race Number of Cases Average Weeks1

Non-black 1771 46.9Black 623 45.0Total 2394 46.4



Table 4.27.2 Number of Weeks Worked/Year by # of Survey Years Reported Work (Unweighted): Older Men

# YearsReported Work

Non-black Black Total# of Cases Average Weeks # of Cases Average Weeks # of Cases Average Weeks

0 22 — 15 — 37 —1–4 74 39.6 39 40.4 113 39.95–8 583 46.9 215 45.4 798 46.5

9–11 606 47.5 235 46.4 841 47.212 486 49.4 119 48.8 605 49.3

Table 4.27.3 Number of Weeks Worked by Survey Year (Unweighted): Older Men

Respondents Working Average Weeks Worked Respondents Not Working Respondents Missing

Year Total Non-black Black Total Non-

black Black Total Non-black Black Total Non-

black Black

1966 4709 3425 1284 48.6 49.2 47.0 311 175 136 — — —1967 4393 3191 1202 49.3 49.7 48.2 351 211 140 276 198 781968 4001 2950 1051 47.5 48.2 45.5 647 388 259 372 262 1101969 3994 2901 1093 48.8 49.1 48.0 387 240 147 639 459 1801971 3764 2751 1013 50.6 50.8 50.2 411 265 146 845 584 2611973 3229 2382 847 47.4 47.6 46.7 722 486 236 1069 732 3371975 2659 1971 688 46.1 46.3 45.6 1073 740 333 1288 889 3991976 2312 1702 610 45.7 46.1 44.6 1175 819 356 1533 1079 4541978 1823 1366 457 44.8 45.3 43.2 1396 973 423 1801 1261 5401980 1444 1090 354 44.2 44.5 43.2 1557 1098 459 2019 1412 6071981 1246 945 301 43.3 43.7 42.1 1586 1132 454 2188 1523 6651983 850 661 189 40.4 40.6 39.7 1783 1272 511 2387 1667 720

Table 4.27.4 Number of Employers by Survey Year (Unweighted): Older Men

Respondents Working Average # of Employers Respondents Not Working Respondents Missing



black Black

1966 4697 3417 1280 1.2 1.2 1.3 323 183 140 — — —1967 4403 3198 1205 1.0 1.0 1.1 341 204 137 276 198 781968 4013 2957 1056 1.2 1.2 1.2 635 381 254 372 262 1101969 3999 2905 1094 1.1 1.1 1.1 382 236 146 639 459 1801971 3802 2774 1028 1.2 1.2 1.3 373 242 131 845 584 2611976 2312 1702 610 1.1 1.1 1.1 1175 819 356 1533 1079 4541981 1248 947 301 1.1 1.1 1.1 1584 1130 454 2188 1523 665

Note: Questions about number of employers missing in some years.



Table 4.27.5 Number of Weeks Worked All Survey Years (Unweighted): Young Men

Race Number of Cases Average WeeksNon-black 1688 44.6

Black 392 43.1

Total 2080 44.3

Table 4.27.6 Number of Weeks Worked Per Year by Number of SurveyYears Reported Work (Unweighted): Young Men

# YearsReported Work

Non-black Black Total# of Cases Average Weeks # of Cases Average Weeks # of Cases Average Weeks

0 5 — 0 — 5 —1–8 46 32.8 24 40.5 70 35.4

9–11 531 41.3 156 39.9 687 41.012 1106 46.7 212 45.8 1318 46.6

Table 4.27.7 Number of Weeks Worked by Survey Year (Unweighted): Young Men

Respondents Working Average Weeks Worked Respondents Not Working Respondents Missing



black Black

1966 4462 3264 1198 33.8 34.6 31.7 763 523 240 — — —1967 4228 3088 1140 36.1 37.1 33.5 562 389 173 435 310 1251968 4037 2961 1076 38.4 39.1 36.3 281 194 87 907 632 2751969 3850 2860 990 39.9 40.6 37.7 183 116 67 1192 811 3811970 3585 2682 903 42.7 42.9 42.1 408 280 128 1232 825 4071971 3721 2809 912 41.8 42.3 40.3 266 189 77 1238 789 4491973 3876 2983 893 46.6 46.8 45.9 138 87 51 1211 717 4941975 3796 2949 847 46.3 47.1 43.5 181 113 68 1248 725 5231976 3326 2599 727 46.8 47.5 44.4 369 249 120 1530 939 5911978 3441 2720 721 48.9 49.4 47.2 97 55 42 1687 1012 6751980 3334 2642 692 48.5 49.0 46.5 104 56 48 1787 1089 6981981 2952 2333 619 49.8 50.0 49.2 446 289 157 1827 1165 662



Table 4.27.8 Number of Employers by Survey Year: Young Men (Unweighted)

Respondents Working Average # of Employers Respondents Not Working Respondents Missing



black Black

1966 4469 3268 1201 1.9 1.9 2.0 756 519 237 — — —1967 4091 2999 1092 1.4 1.4 1.4 699 478 221 435 310 1251968 4045 2966 1079 1.8 1.8 1.8 273 189 84 907 632 2751969 3858 2866 992 1.8 1.8 1.9 175 110 65 1192 811 3811970 3693 2758 935 1.7 1.7 1.8 300 204 96 1232 825 4071971 3790 2858 932 1.6 1.6 1.7 197 140 57 1238 789 4491976 3476 2713 763 1.3 1.3 1.2 219 135 84 1530 939 5911978 3447 2724 723 1.3 1.3 1.3 91 51 40 1687 1012 6751980 3348 2647 701 1.2 1.2 1.2 90 51 39 1787 1089 6981981 3041 2395 646 1.1 1.1 1.1 357 227 130 1827 1165 662


Appendices

201

Appendix A: Quick Reference Guide


This appendix summarizes some key pieces of information regarding this User’s Guide and other documentation products for the Older and Young Men, the NLS surveys, and the NLS CD-ROMs. It also provides answers to some common questions about the Older and Young Men and tells users how to get additional help when necessary.

A.1 Guide to NLS DocumentationThis User’s Guide provides in-depth information about the Older and Young Men data. Users requiring general information about the history, administration, or other cohorts of the NLS project should consult the following table for the most appropriate NLS document.

Table A.1 List of NLS Documentation

NLS Handbookwww.bls.gov/nls/handbook/nlshndbk.htm

This comprehensive introduction to the NLS gives readers general information about all NLS cohorts and the main topics of investigation for each.

NLS Web Homepagewww.bls.gov/nls

This internet site offers an overview of the NLS programs. It also offers downloadable data sets and documentation items for several NLS cohorts.

NLS Internet Bibliography www.nlsbibliography.org

This on-line searchable database provides citations for research using NLS data.

NLS News www.bls.gov/nls/nlsnews.htm

NLS News, a quarterly newsletter, is mailed at no charge to all data users and other interested persons. The newsletter contains informative articles about the NLS data, announces the release of new data, and reports any errors discovered in the data or documentation. Past issue of NLS News are available on the NLS internet site.

NLS User’s Guideswww.bls.gov/nls, follow link for cohort of interest

These cohort-specific guides help researchers understand NLS variables, survey instruments, documentation techniques, and other technical issues.

QuestionnairesThe complete set of survey instruments used with the cohort in each survey year allows researchers to view questions, supplemental information, and household interview forms.

Flowcharts Schematic diagrams depict universe information and skip patterns for many survey instruments.

Codebook Supplements Supplementary attachments and appendices contain variable creation, description, and coding information not present in the questionnaires.

NLS Compact Disc User’s Guides The guide provides installation and usage instructions for the CD-ROMs.

Error updates via the internethttp://www.bls.gov/nls, follow link for cohort of interest

Information about data errors and ways to obtain corrected data is available for each cohort.


http://stats.bls.gov/nlshome.htm


Users may obtain several of the documents listed in Table A.1 on-line from the NLS web site. The remainder of the documents are available by contacting NLS User Services or by going to www.bls.gov/nls and choosing the “Ordering Data” link. Contact information for NLS User Services is provided under “Additional Support” at the end of this appendix.

A.2 About the Older and Young Men CD-ROMThe Older and Young Men survey were both panel data collections—specifically, the same persons were interviewed year after year. Older and Young Men respondents were often asked the same or similar questions in different surveys to gauge the change in behavior over time.

The database is organized by respondent; information on each respondent is stored in a record. In each respondent-specific record, the variables are arranged in chronological sequence. Users are provided with Windows-based data extraction software on each CD-ROM to search this large database and extract the specific variables needed. Researchers should note that, although data for both cohorts are contained on the same CD, they are stored in separate files. Only one cohort’s data can be accessed at a time.

In addition to data on the Older and Young Men, the Original Cohorts CD-ROM also contains information from the Mature and Young Women surveys. These data can be accessed in the same manner as the men’s variables.

Some data and documentation available on the Original Cohorts CD-ROM also are available for download from the NLS web site at www.bls.gov/nls under the “Ordering Data” link.

Search StrategiesVariables can be selected for extraction either using a Search List (reference number or question number) or Search Index (any word in context or year). The following is a brief description of the search options.

Reference Number: Each variable is assigned a reference number that determines its relative chronological position in the database. That reference



number never changes—even when waves of data are added and the database is revised. The reference number is the equivalent of a unique variable name.

Question Number: Each year, a separate survey instrument is used to collect data. A questionnaire item (or question number) refers to the location of a given variable in the printed or electronic questionnaire.

Any Word: The database retrieval software allows the user to search for and select those variables whose titles contain any single word or combination of words.

Year: The user can select a specific survey year and choose variables collected in that specific year.

Accessing the DataThis section briefly describes how to access the data from the Older and Young Men CD-ROM. The reader should consult the DB Investigator Manual for more detailed information about the Windows-based database retrieval system.

Hardware Requirements. The minimum computer standards required to access the Older Men and Young Men data are an IBM-compatible personal computer (PC) running Microsoft Windows 95 or later; a CD-ROM drive connected to the PC; and a writable disk drive connected to the PC.

Users can extract data for specific subsamples and in various formats. Step-by-step instructions for extracting data and detailed information about examining the codebook can be found in the NLS Compact Disc User’s Guide.

A.3 Glossary of NLS Survey TermsAny Word Search. This CD-ROM search function allows users to select any word or words and to view all variables which contain those words in their titles.

Bureau of Labor Statistics (BLS). This agency of the U.S. Department of Labor sponsors and oversees the National Longitudinal Surveys project.

Center for Human Resource Research (CHRR). A research unit at The Ohio State University, CHRR is responsible for the management of the Original Cohorts, documentation and dissemination of the data, and user services.



Children of the NLSY79. This survey group comprises all children born to female NLSY79 respondents. The group was first surveyed in 1986 and has been reinterviewed bienially. Beginning in 1994, a separate survey has been administered to the children age 15 and older, referred to as the “Young Adults.”

Codeblock. Information about each variable is presented in a consistent form called a codeblock. Most codeblocks provide users with the variable title, reference number, question number, survey year, coding information, and a frequency distribution.

Codebook. The codebook contains complete information about all the variables in a data set and is included electronically on the CD-ROM. It comprises a number of codeblocks presenting information about individual variables.

Computer-Assisted Personal Interview (CAPI). These interviews, used since 1993 for the NLSY79 and since 1995 for both women’s cohorts, are administered using a survey instrument on a laptop computer.

Household Record Card. This survey instrument was used during PAPI interviews to collect information about members of the respondent’s household. During the interview, demographic information was transferred from the cards to the “Household Roster” section of the main questionnaire, so data from the Household Record Cards contain “Household Roster” as part of their variable titles.

Household Screener. This survey instrument was used in 1966 to identify respondents eligible for the Older and Young Men cohorts, as well as the women’s cohorts. It collected demographic information about all members of each surveyed household.

Mature Women Cohort. This group of 5,083 respondents, ages 30–44 on April 1, 1967, was first interviewed in 1967 and continues to be surveyed on a regular basis. The Mature Women cohort is one of the four NLS Original Cohorts.

National Longitudinal Survey of Youth 1979 (NLSY79). This group of 12,686 male and female respondents was first interviewed in 1979 and is



currently resurveyed on a biennial basis. Respondents in this cohort were ages 14–21 as of December 31, 1978.

National Longitudinal Survey of Youth 1997 (NLSY97). Respondents ages 12–16 as of December 31, 1996, were selected for inclusion in the newest NLS cohort. The NLSY97 numbers 9,022 respondents; interviews with this cohort are currently conducted every year.

Older Men Cohort. This group of 5,020 respondents, ages 45–59 on April 1, 1966, was first interviewed in 1966 and subsequently surveyed 12 additional times before its discontinuation in 1990. The Older Men cohort is one of the four NLS Original Cohorts.

Original Cohorts. The four cohorts (Older Men, Mature Women, Young Men, and Young Women) selected during the 1966 household screening and first surveyed between 1966 and 1968.

Paper-and-Pencil Interview (PAPI). Traditional paper-and-pencil instruments were used for all surveys of the Older and Young Men.

Primary Sampling Unit (PSU). A Primary Sampling Unit consists of one or more Standard Metropolitan Statistical Areas (SMSAs), counties (or parishes in some states), parts of counties (parishes), or independent cities. PSUs were the basis for sampling Older and Young Men respondents.

Reference Number. A reference number is a unique identifying number beginning with “R,” which is assigned to each variable in the data set. Reference numbers never change after they are assigned to the variables from an interview even as additional information is added to the data set from later surveys.

Young Men Cohort. This group of 5,225 respondents, ages 14–24 on April 1, 1966, was first interviewed in 1966 and subsequently surveyed 11 additional times before its discontinuation in 1981. The Young Men cohort is one of the four NLS Original Cohorts.

Young Women Cohort. This group of 5,159 respondents, ages 14–24 on January 1, 1968, was first interviewed in 1968 and continues to be surveyed on a regular basis. The Young Women cohort is one of the four NLS Original Cohorts.



A.4 How to Get HelpSometimes users have questions about the Older and Young Men data, database retrieval software, or documentation. The following are strategies for finding answers to these questions.

On-line and Paper Documentation Help. NLS User Services provides both on-line and paper documentation help sources for users. Choosing “Help” in the menu bar of the Investigator program leads the user to an on-line manual or a download feature for on-line help, including the NLS of Older and Young Men User’s Guide. Paper documentation help includes the NLS Handbook and assorted supplemental hard copy documentation (see Table A.1).

Frequently Asked Questions. Table A.2 provides answers to commonly asked questions about the Older and Young Men data.

Table A.2 Frequently Asked Questions

Question Answer

Is there a distinction between valid and invalid skips? What is the difference between a –998 and -999 code?

There is no distinction between valid and invalid skips in the data. Noninterviews, valid skips, and invalid skips are lumped together as NAs (not applicable) in the codebook. If you wish to separate valid skips from invalid skips, you must design your own program by using the questionnaires and following the skip patterns. In the data, codes of -128 and -999 are used to represent NAs, depending on the variable width. Codes of -127 and -998 are used to represent DKs (don’t know). More information on the coding of missing responses is presented in section 3.3 of this guide.

Do the sampling weights correct for oversampling?

Yes, the first year weights correct for oversampling. The weights for each subsequent survey year correct for attrition and oversampling.

How can multiple respondent households be identified?

Use the “search any word” feature of the Search Engine to search for all occurrences of the words “identification” and “code.” This search will result in a listing of variables that identify members of the NLS Original Cohorts who are related to the respondent.

How can an employment tenure variable be created?

Tenure with an employer can be constructed with information on start and stop dates of the current/last job and intervening jobs between surveys in conjunction with reports on weeks unemployed or out-of-the-labor force. Consult the “Work Experience” section of this guide for further details and possible limitations of the data.

Is there any interview that provides a geographic residence variable smaller than region?

No. This database does not contain a residence variable smaller than region. The regional distinction is South/non-South. More information about geographic variables is provided in the "Geographic Residence & Environmental Characteristics" section of this guide.

How can users determine who the respondent is for the 1990 survey of Older Men?

The 1990 interview surveyed not only the original Older Men respondents (called “sample persons” in this guide) but also the widows or next-of-kin for deceased respondents. In addition, other proxies such as staff members in an institution were used if the sample person or widow was alive but incapable of completing the survey. To identify whether the respondent is the sample person, widow, next-of-kin, or another proxy, researchers should use R06015.90, “Identification / Residence Status of All 90 Interviews.”

Why is the ‘Total Family Income’ variable not available

‘Total Family Income’ is a KEY variable which is only created for survey years in which a personal interview is conducted. See Table 2.4.1 for a listing of the type of



in all years? interview by survey year.

Additional Support. If questions arise which are not answered in the documentation, contact NLS User Services at:

NLS User ServicesCenter for Human Resource Research

921 Chatham Lane, Suite 100Columbus, Ohio 43221-2418

614-442-7366

E-mail: [email protected]


Appendix B: Original Cohort Dictionary of Acceptable Keywords

Appendix B: Original Cohort Dictionary of Key Words

On the original data tapes, the men’s data were searchable by “keyword”—that is, the user could select from a predetermined list of key common words in variable titles. Although this search option has been replaced by the any word search, the keywords remain in the variable titles. This list of key words used for the men’s cohorts is provided to help researchers focus their any word searches by employing common words included in variable titles for a given topic. Users should note that some key words apply to only one cohort.

*KEY*1st Job2 or More355 Years6 or More MonthsAbsentADAAFDCAgeAge 14Age 15Age 30AlimonyAlternative JobAmount of FatigueAnsweredAnswersApartmentArmed ForcesAssetsAssistanceAttendedAttendingAttitudeAttitudes

AutomobileBenefitsBest JobBirthBirthplaceBondsBooksCensus BureauCensus DivisionCertificateChildChild CareChildrenClass of WorkerCollective BargainingCollegeCollege DegreeCollege SurveyCollegesCommercialCommitment to WorkCompany TrainingComparative Job StatusContactCorrectional InstitutionCounseling


Original Cohort Dictionary of Acceptable Keywords211

CounselorsCultural ExposureCurrent Job XX (For each survey 66–90)Current or Last Dual JobCurrent or Last Job XX (For each survey 66–90)Current SchoolCurriculumDays per WeekDebtDependentsDesireDesiredDifferent AreaDiplomaDisabilityDiscriminationDisliked MostDistrict WideDraftDual JobDual JobsDuncanDuncan IndexEducationEducationalElementaryEmployee AssociationEmployerEmployersEmploymentEmployment Status

Employment Status RecodeEnrolledEnrollmentEnrollment StatusExpectedExpelledExpenditure per PupilExpensesExtra-CurricularFacultyFamilyFamily MemberFamily MembersFarmFatherField of StudyFinancial AidFinancial PositionFood StampsForeign LanguageFull-TimeFunctional LimitationsGEDGED ScoreGoalGrade AttendedGrade AttendingGrade CompletedGraduateGraduatedGrandfatherGrandmotherGuidance


212 Original Cohort Dictionary of Acceptable Keywords

Head of HouseholdHealthHeightHelpingHigh SchoolHigh School SubjectHomeworkHourlyHoursHours per DayHours per WeekHours WorkedHouseHouseholdHousehold ChoresHousehold RecordHouseworkHusbandHypothetical Job OfferIdentification CodeIncomeIndex of DemandIndustryInfluenceInterfereInterfirmInterruptionInterview DateInterview LengthInterview MethodInterview StatusIntrafirmIQ CategoryIQ ScoreIQ Test Name

JobJob before Current Job 66Job CharacteristicsJob FamilyJob LevelJob PaceJob PressuresJob SatisfactionJobsKnowledgeLabor ForceLabor Force StatusLabor MarketLast InterviewLast JobLast Year in High SchoolLayoffLeisureLiabilityLibraryLibrary CardLiked MostLimitLimitationsLiving TogetherLocateLocationLongest JobLoss of Current JobMagazinesMarital StatusMarket ValueMarriageMedicalMedical Care


Original Cohort Dictionary of Acceptable Keywords213

Medical InsuranceMethod of FindingMethod of SeekingMigrant StatusMilitaryMortgagesMost Recent CollegeMost Recent JobMotherMotivationMoveMoved BackwardMutual FundsNationalityNewspapersNoninterviewNonmobile WorkersNot Empld XX (For each survey 66–90)OccupationOccupation DesiredOccupationalOLFOLF XX (For each survey 66–90)On the JobOn the Job TrainingOut of SchoolOvertimeOvertime PayParentsPart-TimePartnerPension

Plan to SeekPlansPoverty StatusProbationProgressedPromotionPropertyPublic AssistanceRaceRankRate of PayReading MaterialReal EstateRentResidenceResponsibilityRestrictionsRetiredRetirementReturn to CollegeReturn to High SchoolReturn to SchoolReturned to SchoolRotter ScaleRound TripSame AreaSampling WeightSavingsSavings BondsSchool QualitySchool SurveySeekSeeking


214 Original Cohort Dictionary of Acceptable Keywords

Self-EmployedSeparationServicesSexSiblingSiblingsSMSASocial SecuritySpanish AmericanSpell Not WorkingSpells of UnemploymentSponsorSSIStocksStudySummerSurvey WeekSurvey Week XX (For each survey 66–90)SuspendedSVP ScoreTeacherTeachersTechnicalTenureTrainingTransportation

TravelTuitionUnemployedUnemployed XX (For each survey 66–90)UnemploymentUnemployment CompensationUnionVeteran BenefitsVeteran CompensationVocationalVolunteerWagesWeeks in Labor ForceWeeks Not WorkingWeeks OLFWeeks UnemployedWeeks WorkedWeightWelfareWifeWomen WorkingWork ScheduleWorkmen CompensationYouth


Appendix C: Unpacking Multiple Entries


Responses to multiple entry questions found in early years of the surveys of the four Original Cohorts were coded in a geometric progression format to conserve space on the tape. Variables such as ‘Method of Seeking Employment,’ ‘Method of Finding Current or Last Job,’ ‘Type of Financial Aid Received,’ and numerous health-related questions have been formatted in this way since the surveys began. Multiple entry items are identified by an asterisk under the source code box in the questionnaire and by a special detailed codeblock in the documentation. These responses need to be “unpacked” before they can be used in analysis. Although the example below pertains specifically to the Young Men’s cohort, it is applicable to the Older Men as well.

Example: Codes for the variable R04561., ‘Fringe Benefits Made Available to R by Current Employer, 1976 (mark all that apply),’ range from 1 (the respondent reported only one such benefit, “medical insurance”) to 259 (the respondent reported “medical insurance,” “life insurance,” and “paid sick leave”) to 1023 (the respondent reported that he had access to all of the benefits listed). Although there are several different ways to sort out which respondent has positive answers on which components, this appendix provides one example in SAS and one example in SPSS.

Program 1: Unpacking Fringe Benefits Data in SASTodd Idson and Bill LeBlanc, in their work on employer size and turnover based on data from the NLS of Young Men, wrote the following SAS program to “unpack” or create dichotomous “yes-no” variables from such fringe benefit data. The program statements listed below can be modified by the user to include the expanded set of fringe benefits available in later survey years as well as to unpack other multiple entry variables by extending the dummy, the counter, and the number of variables to agree with the total number of responses listed in the codeblock in the documentation.

This SAS program unpacks fringe benefits from the variable “fringe.” It creates 10 (dichotomous) dummy variables indicating the presence or absence of each of the 10 benefits. Each dummy is set to missing if fringe is missing (coded at -


999). Note that the variables are created in reverse order from the codeblock, i.e., MEDICAL is code 1 on the tape and FRINGE10 in the program.

TITLE1 ‘FRINGE BENEFITS’;DATA BENEFITS; INFILE IN;INPUT FRINGE 18985-18988;

*TAPE POSITION FOR THE 1976 YOUNG MEN;LABEL FRINGE10 = ‘MEDICAL’

FRINGE9 = ‘LIFE INS’FRINGE8 = ‘RETIRE’FRINGE7 = ‘TRNG’FRINGE6 = ‘PROF SH’FRINGE5 = ‘STOCK’FRINGE4 = ‘FRE MEAL’FRINGE3 = ‘FRE MERC’FRINGE2 = ‘PD SICK’FRINGE1 = ‘PD VAC’;

ARRAY DUMMY(10) FRINGE1-FRINGE10;I=0;DO COUNTER=512, 256, 128, 64, 32, 16, 8, 4, 2, 1;I+1;IF COUNTER <= FRINGE THEN DO;

DUMMY(I) = 1;FRINGE = FRINGE - COUNTER;

END;ELSE DUMMY(I)=0;IF FRINGE = -999 THEN DUMMY(I) = .;

*OMIT TO SET DUMMY=0 IF MV;END; DROP I FRINGE;

PROC MEANS; VAR FRINGE1-FRINGE10;TITLE2 ‘PROPORTION OF (NONMISSING) RESPONDENTS HAVING EACH BENEFIT’;

Program 2: Unpacking Fringe Benefits Data in SPSSThe SPSS program works in the same way as the SAS program. Users of this alternative package can follow this template.

/* UNPACKING 1981 YOUNG MEN FRINGE BENEFITS: SPSS/

compute FB1=0variable labels FB1 ‘81 NONE’compute FB2=0

variable labels FB2 ‘81 FLEX HRS’compute FB3=0variable labels FB3 ‘81 PAID VACATION’


compute FB4=0variable labels FB4 ‘81 PD SICK’compute FB5=0variable labels FB5 ‘81 FR MERCH’compute FB6=0variable labels FB6 ‘81 FR MEALS’compute FB7=0variable labels FB7 ‘81 STOCK’compute FB8=0variable labels FB8 ‘81 PROFT’compute FB9=0variable labels FB9 ‘81 TRED’compute FB10=0variable labels FB10 ‘81 RETR’compute FB11=0variable labels FB11= ‘81 LIFE’compute FB12=0variable labels FB12 ‘81 HLTH’

compute FB81a=FB81variable labels FB81a ‘VARIABLE FOR NONE’

do if (2048 le FB81)compute FB1=1compute FB81=FB81-2048elsecompute FB1=-4end if


do if (512 le FB81)compute FB3=1compute FB81=FB81-512elsecompute FB=-4end if

do if (256 le FB81)compute FB4=1compute FB81=FB81-256elsecompute FB4=-4endif



do if (32 le FB81)compute FB7=1compute FB81=FB81-32else compute FB7=-4end if



do if (4 le FB81)


compute FB10=1compute FB81=FB81-4elsecompute FB10=-4end if

do if (2 le FB81)compute FB11=1compute FB81=FB81-2else

compute FB11=-4end if

do if (1 le FB81)compute FB12=1compute FB81=FB81-1else compute FB12=-4end if

Index

Index

A

Age, 5, 51–57

Aid to Families with Dependent Children

(AFDC). See also Welfare

Alcohol, 7, 58

American College Testing Program (ACT), 59

Any word search, 49, 166, 171

Aptitude and Intelligence, 59

Armed Forces, 146, 147

Assets, 8, 46, 99–104

Attitudes, 7, 113, 137

Attrition, 9, 24

B

Benefits. See Fringe Benefits

Birth date. See Date of birth

Birthplace, 75

Bradburn Affect Balance Scale, 86

Bureau of Labor Statistics (BLS), 5

C

California Test of Mental Maturity, 59

CD-ROM, 9, 33, 35, 39, 41, 47, 49, 50, 166

Census Bureau, 5

Center for Human Resource Research

(CHRR), 5

CES-D Scale, 86, 87

Children, 124–26

Children of the NLSY79, 4, 167

Cigarette use, 7, 58

Class of worker, 61–62, 107, 109

Codeblock, 41–43, 45

Codebook, 41, 49

Codebook Supplement, 46, 69, 76, 114, 131,

165

College, 65, 66, 90–91

Created variables, 36–37, 41, 45, 104, 147

Crime, 63

Current Population Survey (CPS), 47, 48, 109,

116, 147

D

Data

Sets, 9

Date of birth, 53, 56, 57

Death, 53, 55, 84

Delinquency, 63

Disability, 129, 133, 134, 138

Discrimination, 7, 64

Documentation, 33, 41, 165

Duncan Socioeconomic Index, 47, 48, 131

E

Education, 7, 64–69

Employers

Number, 157, 161. See also Jobs

Employment Status Recode (ESR), 40, 47, 48,

110, 117, 118, 119

Enumeration districts (EDs), 14

Environmental characteristics, 8, 75

Error Updates, 49

Index

Ethnicity, 142–43

F

Family

Background, 70–72. See also Household-

Composition

Fertility, 8, 122–27

Field Representative’s Manual. See

Interviewer's Reference Manual

Fielding periods, 9, 15, 16, 41

Fielding procedures, 18

Flowcharts, 37

Food stamps, 140, 141

Fringe benefits, 73, 129

G

General Education Development (GED), 131

Geographic mobility, 75

Geographic regions, 8

H

Hardware requirements, 167

Health, 8, 80–89

Health insurance, 84

Height, 87

Help, 169–70

High School, 65, 66, 90–91

Household

Composition, 8, 92–98

Members, 92

Household Record, 48

Household Record Card, 17, 33–36, 92, 127,

142

Household Record Variables, 47

Household responsibilities, 127

Household Roster, 53, 71

Household Screener, 34

I

Income, 8, 99–104, 129

Industry, 105–10

Information Sheet, 35, 36, 41, 127

Interviewer’s Reference Manual, 20, 21, 33,

34, 36, 118

J

Job Characteristics Index (JCI), 111

Job Diagnostic Survey (JDS), 111

Jobs

Characteristics, 131. See also Employers

Hypothetical offers, 114, 150

Satisfaction, 112–15

K

KEY variables. See Created variables

L

Labor force status, 8, 106, 116–22

Leisure activities, 8, 138

Lorge-Thorndike Intelligence Test, 59

M

Main data elements, 7

Marriage, 8, 122–27

Status, 123, 125

Mature Women cohort, 4, 5, 40, 114, 127, 166

Medicaid, 84

Medicare, 84, 86

Mental Health, 85

Military, 9, 48, 128–29

Missing Responses, 45

Index

Multiple respondent households, 7, 14, 95, 96,

102

Multiple response questions, 44

N

National Institute of Child Health and Human

Development (NICHD), 4

National Longitudinal Survey of Youth 1979

(NLSY79), 4, 5, 168

National Longitudinal Survey of Youth 1997

(NLSY97), 4, 5, 168

National Longitudinal Surveys (NLS), 3, 6

National Opinion Research Center (NORC), 6

Nationality, 142–43

NLS News, 49

NLSY79, 114

NLSY79 Children, 5

NLSY79 Young Adults, 4, 5

Noninterview, reason for, 18, 20, 21, 22, 23

Nonresponse, 28, 45, 102

O

Occupation, 109, 129–32

Coding, 47, 48

Prestige, 47, 48, 131

Occupation & Industry (O & I) Rewrite, 105,

107, 109

Older Men 1990 resurvey, 18, 19, 20, 34, 38,

43, 46, 47, 53, 55, 58, 82, 84, 86, 93, 102,

109, 113, 116, 119, 123, 124, 134, 144, 152,

155, 157

Older Men cohort, 3, 5, 35, 40, 47, 57, 92, 140

Original Cohorts, 4, 9, 24, 34, 40, 44, 57, 72,

74, 75, 97, 166

Otis/Beta/Gamma, 59

P

Paper-and-pencil instruments (PAPI), 6, 36, 45

Parents, 71

Partner, 92, 124, 125

Pensions, 133–39

Primary Sampling Unit (PSU), 13, 168

PSAT, 59

Public Assistance, 140–41

Public housing, 141

Q

Question numbers, 35, 50, 166

Questionnaire, 34, 35

Questionnaire item, 46

R

Race, 22, 23, 24–26, 27, 55, 121, 122, 142–43,

145, 157

Railroad Retirement. See Social Security

Reference numbers, 37, 38, 41, 50, 166, 168

Reference period, 39

Regression analysis, 29

Reservation wages. See Jobs-Hypothetical

offers

Residence, 75–80

Retention rate, 16, 24

Retirement, 9, 73, 133–39

Rotter Internal-External Control Scale, 85, 87,

88

S

Sample

Design, 9, 13, 14, 28

Exclusion, 18

Representativeness, 9, 24

Index

Size, 5, 9, 16

Sample person, 38

Sampling weights, 9, 27, 28, 29

SAT, 59

School discipline, 63

School Survey, 36, 59, 63, 90

Screening process, 13

Search strategies, 166

Short Portable Mental Status Questionnaire

(SPMSQ), 19, 86

Siblings, 71

Social Security, 133–39

Specific Vocational Preparation (SVP), 131

Spouse, 125. See Wife

Standard Metropolitan Statistical Area

(SMSA), 8

Supplemental Security Income (SSI), 140, 141

Survey

Design, 18

Instruments, 33

T

Technical Review Committee, 6

Topcoding, 46, 102

Topical sections, 10

Training, 9, 128, 129, 136, 144–48

U

Unemployment, 140

Universe, 17, 46

User Services, iii, 37, 170

V

Valid values, 46

Variables. See also Created variables

Titles, 38, 39

Veterans benefits, 138

Volunteer work, 8

W

Wages, 149–50

Wealth, 99

Weeks worked, 158, 160

Weight, 87

Welfare, 140, 141

Widow, 38, 94

Wife, 123, 125, 136

Work experience, 151–61

Workers’ Compensation, 140

Y

Year index, 50

Young Men Codebook Supplement, 90

Young Men cohort, 3, 5, 35, 36, 48, 57

Young Women cohort, 4, 5, 36, 63, 66, 90, 91,

114, 127, 166

Index

Documents

1 data/final draft M… · Web viewEach Household Record Card (1) enumerated all persons currently living in the household; (2) recorded for each person: name, relationship to respondent,