THE IMPACT OF MARRIED WOMEN. AND
RELATED FACTORS ON THE
DISTRIBUTION. OF FAMILY EARNINGS

Dissertation for the Degree of Ph. D.
MICHIGAN STATE UNIVERSITY
ROBIN LYNN BARTLETT
19-74

9 .n.»c....’.m'

This is to certify that the

thesis entitled

THE IMPACT OF MARRIED WOMEN
AND RELATED FACTORS ON THE
DISTRIBUTION OF FAMILY EARNINGS

presented by

Robin Lynn Bartlett

has been accepted towards fulfillment
of the requirements for

Ph. D. degree in Economics

Major professor

 

Date February 22, 1974

 

0-7639

 

  

‘7.:‘:‘
“Iran"mfgsfirg

   

I BUOI BINDFRY INC

9 r as r (was
III SPRINGPORI Amman

  
 

 

 

 

 

ABSTRACT

THE IMPACT OF MARRIED WOMEN AND RELATED
FACTORS ON THE DISTRIBUTION OF FAMILY EARNINGS

BY

Robin Lynn Bartlett

The purpose of this study was to examine some of the factors that
affect the distribution of aggregate family earnings; that is, whether
certain factors make the distribution of aggregate family earnings more
or less disperse. Amang the factors that were suggestions and extensions
of the current literature were the marital labor force participation rates,
relative hourly earnings between all females and males in the labor force,
the relative number of hours supplied between males and females in the
labor force, relative hourly earnings between wives and husbands, and the
relative number of hours supplied between husbands and wives.

In order to construct a family earnings distribution, and to avoid
the heroic assumptions that would have to be made to bridge the gap be-
tween a family earnings distribution and a family income distribution,
the core population of all husband and wife families was chosen. The
necessary data were taken from various published and unpublished sources
of information gained from the Current Population Survey. The data were
in three major groupings: (l) marital labor force participation pairings --
that is, all the possible combinations of husband and wife occupational

pairings if they were employed as well as unemployed or not in the labor

(7‘56? Robin Lynn Bartlett

fbrce employment statuses; (2) hourly earnings and hours worked by sex
and occupation; and (3) the marital labor force participation pairings
as in (1). The time period considered was between 1962 and 1971.

Using Gini coefficients to determine whether or not the factors men-
tioned above had an equalizing or disequalizing impact on the distribution
of family earnings, the fellowing results were obtained. First, the in-
creased labor force participation rates of married women as reflected in
the increased marital labor force participation rates had a substantial
disequalizing impact on the distribution of family earnings. This dis-
equa1izing impact was somewhat offset by the equalizing influences of
the other four factors. The increase in the relative hourly earnings
between all females and males had the greatest equalizing impact of the
four. The net result was the fact that relative income dispersion in-

creased between 1962 and 1971.

THE IMPACT OF MARRIED WOMEN AND RELATED
FACTORS ON THE DISTRIBUTION OF FAMILY EARNINGS

BY

Robin Lynn Bartlett

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Economics

1974

Copyright by

Robin L. Bartlett

1974

TO THE BARTLETT CLAN:

C.D.
MARCELLA
DAN
LINDA
BILL
SHARON
MIKE
BECKY
ROY
BERNIE
AND GRAM

ACKNOWLEDGMENTS

I wish to thank the following people:

Professor Robert Rasche, Chairman of the Committee, for crystalizing
the initial ideas and for patiently reading the numerous drafts of each
chapter;

Professor Collette Moser and The Center for Rural Manpower and Public
Affairs for financial support;

Professor Maurice Weinrobe for his help and supervision;

Professor Daniel Hamermesh for willingness to speedily read and
critique various drafts;

Professor William Haley for his assistance with various statistical
problems;

Various friends who had their own special input;

Jacque Grossman who programmed a substantial portion of the work;

Linda Peppard who read countless numbers;

Jean Carlson who edited critical sections;

Jeanette Barbour and Betty Kronemeyer who typed the first and last
drafts;

Pam Japinga who made the coffee breaks enjoyable and gave support
before and during the writing of this thesis;

Those in my office -- Robert Leekley, Evelyn and Mark Fallek, and

Jan Palmer.

iii

TABLE OF CONTENTS

LIST OF TABLES I O O O O 0 C O O O O O O O O O O O O O O O O O O O 0

LIST OF FIGURES O O O O O O O O O O O O O O O O O O O O O O I O O 0

CHAPTER 2

LUMP-J
000

4.0

CHAPTER 3

NH
00

Introduction. . . . . . . . . . . . . . . . . . . . . . .
Some Suggestions From the Literature. . . . . . . . . . .
The Model . . . . . . . . . . . . . . . . . . . . . . . .
The Outline of This Study . . . . . . . . . . . . . . . .

RELATIVE HOURLY EARNINGS AND HOURS WORKED BETWEEN ALL
MALES AND FEMALES IN THE LABOR FORCE . . . . . . . . . .

Introduction. . . . . . . . . . . . . . . . . . . . . . .
Aggregate Earnings. . . . . . . . . . . . . . . . . . . .
The Data-Sources, Comparability, an Alterations. . . . .
3.1 The CPS. . . . . . . . . . . . . . . . . . . . . . .
3.2 Sources. . . . . . . . . . . . . . . . . . . . . . .
3.3 Alterations. . . . . . . . . . . . . . . . . . . . .
Summary . . . . . . . . . . . . . . . . . . . . . . . . .

AGGREGATE TRENDS AND THE IMPORTANCE OF RELATIVE CHANGES
I N THE LABOR FORCE O O C O O C I O O O O O O O O O O I 0

Introduction. . . . . . . . . . . . . . . . . . . . . . .
Aggregate Trends in Earnings, Hours, and Wages. . . . . .
2.1 Earnings and Wage or Salary Income . . . . . . . . .
2.2 The Relative Quantity of Labor Supplied. . . . . . .
2.2.a The Quantity of Labor Supplied by Occupations
2.2.b The Relative Quantity of Labor Supplied by
Occupations . . . . . . . . . . . . . . . . .
2.2 c The Quantity of Labor Supplied by Industries.
2.2.d The Relative Quantity of Labor Supplied by
Industry. . . . . . . . . . . . . . . . . . .
2.3 Hourly Earnings and Wage Rates . . . . . . . . . . .
2.3.a Hourly Earnings Rates by Occupation . . . . .
2.3.b Relative Hourly Earnings by Occupation. . . .
2.3.c Wage Rates by Industry. . . . . . . . . . . .
2.3.d Relative Wage Rates by Industry . . . . . . .
2.4 Women's Bureau Data and Their Updates. . . . . . . .

iv

vi

I‘\JUlI'-‘I-J

14

14
14
23
23
24
28
3O

31

31
31
31
33
34

39
42

45
47
48
50
5]
51
52

3.0 The Importance of Relative Quantities of Labor Supplied
and Relative Remuneration Rates Between Females and
Males . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.0 Occupational Shifts in the Labor Force. . . . . . . . . . 58
4.1 Shifts in Occupations. . . . . . . . . . . . . . . . 58
4.2 The Impact of Occupational Shifts. . . . . . . . . . 60
5.0 Summary and Conclusions . . . . . . . . . . . . . . . . . 62

CHAPTER 4 AGGREGATE TRENDS AND THE IMPORTANCE OF RELATIVE HOURLY
EARNINGS AND HOURS WORKED BETWEEN HUSBANDS AND WIVES IN
TIE ECONOMY O I O O O C C C C O C O C . C O C O O C C O O 64

1 0 Introduction. . . . . . . . . . . . . . . . . . . . . . . 64
2.0 Aggregate Earnings of All Husband and Wife Families . . . 64
3 0 The Data - Sources and Alterations. . . . . . . . . . . . 69
3.1 Sources. . . . . . . . . . . . . . . . . . . . . . . 69
3.2 Alterations. . . . . . . . . . . . . . . . . . . . . 69
4.0 Relative Hourly Earnings and Hours Worked Between All

Husbands and Wives in the Economy . . . . . . . . . . . . 70
5.0 A Conditional Probability Distribution of Husbands and

Wives . . . . . . . . . . . . . . . . . . . . . . . . . . 76
5.0 Summary and Conclusions . . . . . . . . . . . . . . . . . 81

CHAPTER 5 THE CONTRIBUTIONS OF MARRIED WOMEN TO AND THE EFFECT OF
RELATED FACTORS ON THE DISTRIBUTION OF AGGREGATE FAMILY
EAm INGS . C C O I O O O O O O C O O O O O O O O O O O O 8 3

Introduction. . . . . . . . . . . . . . . . . . . . . . . 83
Aggregate Family Earnings and Its Distribution. . . . . . 85
The Impact of Related Factors on the Distribution of
Family Earnings . . . . . . . . . . . . . . . . . . . . . 95
3.1 Adjusted Gini Coefficients for Components of

Aggregate Family Earnings. . . . . . . . . . . . . . 95
3.2 Adjusted Gini Coefficients for Related Factors . . . 98

(UNI-J
0
000

CHAPTER 6 A SUMMARY OF FINDINGS AND CONCLUSIONS WITH SUGGESTIONS
FOR POLICY AND FUTURE RESEARCH . . . . . . . . . . . . . 101

A Summary of Findings and Conclusions . . . . . . . . . . 101
Policy Implications . . . . . . . . . . . . . . . . . . . 106
Further Research. . . . . . . . . . . . . . . . . . . . . 106

b) N I-‘
O
O O 0

APPENDIX A. O O O O O O O O O I O O O O O O O O O O O O O O O O O O 108
APPENDIX B. O O O O 0 O O O O O O O O O O O O O O O O O O O O O O O 132

LIST OF REFERENCES. 0 O O O O O O O O O O O O O O O 0 O O O O O O O 134

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

10

11

12

13

LIST OF TABLES

Family Earner Patterns for All Husband and Wife Families in 1971
A Comparison of Median and Mean Incomes

The Ratio of Aggregate Female Earnings to That of Males and the
Ratio of Aggregate Wage or Salary Income of Females to That of
Males, EFt/EMt, 1962 to 1971

The Relative Number of Annual Hours Supplied by All Males to
That Supplied by All Females From Occupational and Industrial

Categories, Hm/HE, 1962 to 1971

Average Annual Percentage Changes in

pation, 1962 to 1971

Selected Variables by Occu-

Average Annual in Selected Relative Variables

By Occupation,

Percentage Changes
1962 to 1971

Average Annual in
Industry, 1962

Percentage Changes
to 1971

Selected Variables by

Average Annual Percentage Changes in Selected Relative Variables

by Industry, 1962 to 1971

The Ratio of Hourly Annual Earnings or Wage Rates Between Females
and Males, Wf/Wfi, Using Aggregate Persons' Data, 1962 to 1971

The Ratio of Aggregate Female Earnings to That of Males and the
Ratio of Aggregate Wage or Salary Income of Females to That of
Males, EFt/EMt, Using Weighted Averages, 1962 to 1971

of
1962

The Relative Aggregate Median Annual Wage or Salary Income
Full-Time, Year-Round Females and Males, WSFTE/WSFTS, From
to 1971

of
Occu-

The Relative Aggregate Annual Median Wage or Salary Income
All Wbrkers Between Females and Males (WSE/WSE) Across All
pations From 1962 to 1971

The Relative Aggregate Median Annual Wage or Salary Income of
Full-Time, Year-Round Workers (WSFTE/WSFTE) andlnl Workers
(WSE/WSE) For All Females and Males, Industrial Data, 1962
1971.

to

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

The Relative Importance of the Variance in the Percentage Change
of Hfi/HE, WE/Wfi, and the Cov (Hg/HE, WE/Wfi) in Explaining the Vari—
ance of EF /EMt, From 1962 to 1971

Laspeyres and Paasche Indices of Hourly Remunerations, Annual
Remunerations by Sex for Occupations and Industries From 1962
to 1971

A weighted Average of Relative Annual and Hourly Remunerations
Between Females and Males for All Occupations and Industries,
From 1962 to 1971

The Ratio of
EWt/EHt , For

Aggregate Wives' Earnings to That of Their Husbands',
A11 Husband and Wife Families, From 1962 to 1971

The Ratio of Aggregate Wives' Earnings to That of Their Husbands',
EWt/EHt, for Families in Which both Husband and Wife work, From
1962 to 1971

The Relative Number of Hours Supplied by Husbands to That Supplied
by Wives, Hfi/Hfi, for All Husband and Wife Families and for Those
in Which Both WOrked, From 1962 to 1971

Relative Hourly Earnings Between Wives and Husbands, WS/Wfi, for
All Husband and Wife Families and for Those in Which Both WOrked,
From 1962 to 1971

The Relative Importance of the Variance in the Percentage Change
of (Hg/HS), (wqé/wfi), and the mung/Hg, wg/wg) in Explaining the
Variance of EW /EHt, From 1962 to 1971

Laspeyres and Paasche Hourly Earnings Indices for All Wives and
All Husbands, From 1962 to 1971

Weighted Averages of the Relative Hourly Earnings Between All
Wives and Husbands, From 1962 to 1971

The Occupational Distribution of All Females, A11 Working Wives
A11 Males and All WOrking Husbands In 1971 By Percentage of
Totals in Each

The Conditional Probability Distribution of All Husband and
Wife Families, 1962 to 1971

Aggregate Family Earnings, AEFAMt, and Husband and Wife Families
in Thousands, From 1962 to 1971

. . . t
Gini Coeffic1ents for Aggregate Family Earnings, AEFAM , From
1962 to 1971

Aggregate Family Earnings and Hours Worked by Quintile, In 1962

and 1971

vii

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

29

30

31

32

33

The Percentage of Wives in Particular Occupations Within Each
Quintile in 1962 and 1971

The Actual Gini Coefficients for Aggregate Family Earnings,
AEFAMt; For All Husbands' Earnings, EH and For All Wives'
Earnings, Ew ; From 1962 to 1971

. . t
The Gini Coefficient for Aggregate Family Earnings, AEFAM , the
Adjusted Gini Coeff'cients for Constant Marital Labor Force Parti-

cipation Rates, IjN ; for Constant Hours worked Per Year, H; and
for Cbnstant Hourly Earnings, fi; From 1962 to 1971

. . t;
The Gini Coefficient for Aggregate Family Earnings, AEFAM the

Adjusted Gini Coefficients for Constant Hourly Earnings of Wives,
WS; and for Constant Hours WOrked Per Year by Wives, Hw; From
1962 to 1971

t;
The Gini Coefficient for Aggregate Familyt Earnings, AEFAM tthe
Adjusted Gini Coefficients Holding iw h/iwt f, iHh/in' jW’vtI/jw h'

t t
2 t 1971
and ij/th From 196 0

THE AVERAGE NUMBER OF WEEKS WORKED PER YEAR PER PERSON BY
OCCUPATION AND SEX

Constant,

THE AVERAGE NUMBER OF WEEKS WORKED PER YEAR PER PERSON BY
OCCUPATION AND SEX, FROM 1962 TO 1971

THE AVERAGE NUMBER OF HOURS WORKED PER YEAR PER PERSON BY
OCCUPATION AND SEX, FROM 1962 TO 1971

THE NUMBER OF WORKERS EMPLOYED BY OCCUPATION AND SEX, FROM
1962 TO 1971, IN THOUSANDS

THE TOTAL NUMBER OF HOURS WORKED ANNUALLY BY SEX AND OCCUPATION,
FROM 1962 TO 1971, IN THOUSANDS
THE HOURLY EARNINGS RATE BY OCCUPATION AND SEX, FROM 1962 TO 1971

THE AVERAGE NUMBER OF HOURS WORKED PER WEEK PER FEMALE RELATIVE
TO THAT PER MALE BY OCCUPATION, FROM 1962 TO 1971

THE AVERAGE NUMBER OF WEEKS WORKED PER YEAR PER FEMALE RELATIVE
TO THAT PER MALE BY OCCUPATION, FROM 1962 TO 1971

THE AVERAGE NUMBER OF HOURS WORKED PER YEAR PER FEMALE RELATIVE
TO THAT PER MALE BY OCCUPATION, FROM 1962 TO 1971

THE NUMBER OF FEMALE RELATIVE TO DULE WORKERS EMPLOYED BY OCCU-
PATION, FROM 1962 TO 1971

THE TOTAL NUMBER OF HOURS WORKED ANNUALLY BY ALL FEMALES RELA-
TIVE TO THAT OF ALL MALES BY OCCUPATION, FROM 1962 TO 1971

viii

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

TABLE

THE HOURLY EARNINGS RATE PER FEMALE RELATIVE TO THAT PER MALE
BY OCCUPATION, FROM 1962 TO 1971

THE MEDIAN ANNUAL WAGE OR SALARY INCOME PER FEMALE RELATIVE TO
THAT PER MALE BY OCCUPATION, WOMEN'S BUREAU UPDATE, FROM 1962
TO 1971

THE AVERAGE NUMBER OF HOURS AT WORK PER WEEK PER PERSON BY
INDUSTRY AND SEX, FROM 1962 TO 1971

THE AVERAGE NUMBER OF WEEKS WORKED PER YEAR PER PERSON BY
INDUSTRY AND SEX, FROM 1962 TO 1971

THE AVERAGE NUMBER OF HOURS WORKED PER YEAR PER PERSON BY
INDUSTRY AND SEX, FROM 1962 TO 1971

THE NUMBER OF WORKERS EMPLOYED BY INDUSTRY AND SEX, FROM 1962
TO 1971, IN THOUSANDS

THE TOTAL NUMBER OF HOURS WORKED ANNUALLY BY SEX AND INDUSTRY,
FROM 1962 TO 1971, IN THOUSANDS

THE HOURLY WAGE RATE BY INDUSTRY AND SEX, FROM 1962 TO 1971

THE AVERAGE NUMBER OF HOURS WORKED PER WEEK PER FEMALE RELATIVE
TO THAT PER MALE BY INDUSTRY, FROM 1962 TO 1971

THE AVERAGE NUMBER OF WEEKS WORKED PER YEAR PER FEMALE RELATIVE
TO THAT PER MALE BY INDUSTRY, FROM 1962 TO 1971

THE AVERAGE NUMBER OF HOURS WORKED PER YEAR PER FEMALE RELATIVE
TO THAT PER MALE BY INDUSTRY, FROM 1962 TO 1971

THE NUMBER OF FEMALE RELATIVE TO MALE WORKERS EMPLOYED BY
INDUSTRY, FROM 1962 TO 1971

THE TOTAL NUMBER OF HOURS WORKED ANNUALLY BY ALL FEMALES RELA:
TIVE TO THAT OF ALL MALES BY INDUSTRY, FROM 1962 TO 1971

THE HOURLY WAGE RATE PER FEMALE RELATIVE TO THAT PER MALE BY
INDUSTRY, FROM 1962 TO 1971

THE MEDIAN ANNUAL WAGE OR SALARY INCOME PER FEMALE RELATIVE TO

THAT PER MALE BY INDUSTRY, WOMEN'S BUREAU UPDATE, FROM 1962 TO
1971

ix

LIST OF FIGURES

FIGURE 1 A Hypothetical Lorenz Curve for Aggregate Family Earnings

FIGURE 2 The Lorenz Curves or Aggregate Family Earnings in Selected Years

CHAPTER 1

THE DISTRIBUTION OF FAMILY EARNINGS: AN OVERVIEW

l..0 Introduction

The purpose of this study is to examine some of the factors that
‘affect the distribution of family earnings; that is, whether certain factors
Hake the distribution of family earnings more or less disperse. Particu-
lar attention is given to working wives and their contributions to aggre-
gate family earnings and its distribution. This special attention is
warranted and of interest for two reasons. First, the increased labor
force participation of married women has many major economic and social
ramifications, one of which is its impact upon the distribution of aggre-
gate family earnings. Second, the work that has already been done in
this area is inconclusive. Many studies pertaining to income size dis-
tribution and to the labor force participation of married women have made
valuable suggestions as to which factors should be taken into consider-
ation. After examining a few of these suggestions, an attempt will be

made to integrate them into a model.

2.0 Some Suggestions From the Literature

From the macro perspective of income size distribution studies and
the micro perspective of labor force participation studies, the consensus
is that the increased labor force participation of married women is having

and will continue to have an impact on the distribution of family earnings.

The direction and magnitude of this impact, however, depends upon the
muIEancing or retarding influences of related factors traditionally con-
sidered by each perspective. The following selections from the literature
foous upon these related factors.

To Herman Miller, the most important factor relating to the increased
laibor force participation of married women is the proportion of families
vwith working wives in the total population of all families. The total
lpopulation of all families is divided into four groups: (1) Male head,
Inarried, husband present, and wife not in the paid labor force; (2) Male
head, married, husband present, and wife in the paid labor force; (3) Male
head, other; and (4) Female head. Miller looks at the respective Gini
coefficients of the first two groups and infers that:

"...the incomes are much more equally distributed among

families where the wife works than where she is not working;

the sizeable increase in thegroportion of families with work-

ing wives has tended to decrease income inequality during

the past decade." [13, Miller, p. 22]

Despite Miller's reference to the 1950's, he would probably make the same
inference with respect to the 1960's since the proportion of families
with a working wife has continued to increase. Miller is implicitly assum-
ing that the shapes of the individual income distributions for these two
groups of families are not changing as a result of the increased labor
force participation of married women.

In a more recent study, Charles Metcalf, initially agreeing with
Miller, shows that in fact the underlying distributions do change and
that the impact of working wives cannot be readily surmized.

"The increase in the proportion of families with a wife in

the labor force has an impact on the aggregate distribution

for two reasons. First, they are shaped differently, a change

in their relative proportion will appear as a change in the
aggregate distribution even if there is no change in the

shapes of the two individual distributions. Second, the two

groups appear to show different responses to cyclical move-

ments in the economy." [4, Metcalf, p. 66]

The: cyclical movements, moreover, of the two underlying distributions
wezne found by Metcalf to vary inversely. He found a positive correlation
Ixrtween his measure of relative income equality and cyclical movements
:fi): the distribution of families with male head, married, husband present,
aund wife in the paid labor force. Counter-cyclical behavior was recorded
‘fior those families with Male head, married, husband present, and wife not
in the paid labor force. The two studies point out the necessity of in-
‘vestigating the proportion of wives that work.

From a different perspective, the cyclical labor force behavior of
husbands and/or their wives has been documented in several labor force
participation studies. A few of the major ones are those done by Bowen
and Finegan, Cain, and Mincer. Labor force entry, exit, and number of
hours worked are explained using the traditional notions of income and
substitution effects. The question as to why a wife works, however, is
not as relevant to this study as how much she works and for what wage.

In the unlikely situation that the "whys" for working between husband

and wife are independent, George Stigler hypothesizes that "...the effect
of their work [wives] would be to increase, and very substantially, the
dispersion of family earnings." [1, Stigler in Budd, p. 104] Edward Budd
is quick to point out that this is only true if a special case is made for

this unlikely situation:

"It does not follow, for example, that the distribution of
family income in a world in which some wives as well as hus-
bands work is necessarily more unequal than the family income
distribution in a world in which only husbands work and wives
stay at home, even if in the first world a wife's decision to
work is independent of whether her husband does so. Only if

there is a very close and direct relationship between the

amount earned by a working wife and the corresponding size

of her husband's earnings would the degree of income equality

of family income be greater in the first case than in the

second." [1, Budd, p. xxvii]
As. an example, in a world where substitution and income effects do not
erlter into the decision of a wife to work, the increased labor force par-
tuicipation of married women would have a greater disequalizing effect if
time relative earnings of wives were 90 percent of their husbands', than
say 40 percent .

However, Budd's special case of Stigler's unlikely situation may be
a crucial element of Jacob Mincer's more general situation. Lodged within
Mincer's permanent and transitory income explanation of the labor force

participation of married women, he notes:

"For particular population groups observed, the more prevalent
transitory components in the head's income, and the weaker the
association between the wage rate of family members, the greater
the equalizing effect of labor supply on family income distri-
bution--and conversely." [5, Mincer, p. 97] (Emphasis added)

 

Or in other words, where transitory components in the head's income are
less prevalent, a weaker association between the husband's wage and that
of his wife would have less of a disequalizing impact upon the distribu-
tion of family income than a strong association. If transitory components
are prevalent in the head's income the opposite would be true. The extent
of transitory components in most head's income, and thus family income,
dampens the disequalizing of family income resulting from the increased
labor force participation of wives. Mincer's quote could also be inter-
preted as emphasizing the buffering influence of the relative wage between
husband and wife upon the prevalence of transitory or permanent income
components of the head's income. For the purposes of a family income

distribution, the former is more adequate. It does mean that if there

is j_n fact a high correlation between the husband's and wife's wage, an
nuzxrease in the number of working wives would make the distribution of
family income more unequal, but deceptively so in degree.

The literature suggests that the increased labor force participation
Of 'wives could make the distribution of family income more or less equal
depending upon the proportion of families whose wives work and the rela-
tzive wage rate between husbands and wives. Taking the former related fac-
‘tor first, the increased labor force participation of wives could have
\rarying impacts if a larger proportion of high income wives as compared
to low income wives worked. Secondly, a wife is also part of the female
labor force. Her occupational location and hourly wage rate is determined
by those forces experienced by all females. The same is true for her
husband. It stands to reason, therefore, that the relative wage rate be-
tween husbands and wives is related to the occupational distribution of
all males and females and their relative wage rates. If females generally
receive lower wages than males in most occupations, in addition to the
likelihood of being in relatively lower paying occupations; then the rela-
tive wage between husband and wife is smaller, weakening their wage rate
association. Using Mincer's argument, the impact of more working wives

would be to decrease the inequality of family income.

3.0 The Model

While the previous studies dealt generally with all families, this
study will concentrate upon all husband and wife families in the U.S.
The scope is further narrowed since earnings, not total money income,
will be discussed. This is done for two reasons. First, most families

are husband and wife families, and earnings are their major source of

ihcc>rma. Second, establishing all husband and wife families and their
earrlings as a core population avoids the heroic assumptions that would
havee to be made in order to bridge the gap between a family earning dis-
trijoution and a family income distribution.

This core population of all husband and wife families is divided
ixrto four groups:

1. Families in which the husband works and his wife stays at home;

2. Families in which both husband and wife work;

3. Families in which the husband stays at home and his wife works;

and

4. Families in which neither the husband nor his wife works.
The data for 1971 indicate,contrary to popular belief, that the first of
these four groups is not the predominant one. This is determined by re-
grouping the ten possible family earner patterns in Table l in the following
manner. The largest single family earner pattern component of the first
group is that where the husband is the sole provider (2.a.). 15,449
thousand families or 34 percent of all husband and wife families exhibited
this pattern. When those families in which the husband worked along with
some other relative or relatives other than his wife (3.b. + 4.b.) are
added, the total number of all husband and wife families in group 1 is
20,094 thousand or 44 percent of the population. Combining the family
earner patterns in which both husband and wife worked (3.a. + 4.a.) for
group 2 gives 20,552 thousand or 45 percent of all husband and wife families.
Group 2 then is slightly larger than group 1, making it the predominant
family earner group. The remaining family earner patterns in Table l are
partitioned into groups 3 and 4. The former (2.b. + 3.c. + 4.c.) accounts

for only 3 percent of all husband and wife families, families where just

TABLE 1

Family Earner Patterns for All
Husband and Wife Families in 1971*

FAMILY EARNER PATTERN

 

1. No Earners

 

2. One Earner

 

a. Head Only
b. Wife Only
c. Other Relative Only

3. Two Earners

 

a. Head and Wife
b. Head and Other Relative
c. Head Not an Earner

4. Three Earners

 

a. Head and Wife
b. Head an Earner, Wife Not an Earner
c. Head Not an Earner

5. Total Number of Families

 

*[12, 0.8. Department of Commerce, p. 80]

3,443
16,784

18,714

6,812

45,753

15,449
997
338

15,381
3,103
230

5,171
1,542
99

the wife worked. In 8 percent of all husband and wife families, however,
neither the husband nor his wife worked (1. + 2.c.).
The four groupings are easily expressed in algebraic terms.
t t t

= +
EFAM Eh BW (1.1)

t . . . t
where EFAM represents the family's total annual earnings in year t; Eh

. . t .
represents the husband's earnings in year t; and Ew represents the wife's

earnings in year t. In Equation (1.1), Et > 0 and E: = O for those families

h

in which the husband works and his wife stays at home. E: > O and Et > 0

w
for those families in which both work. E: = 0 and E: > O for those families
in which the husband stays at home and the wife works. Finally, E: = 0

t
and Ew = 0 for those families in which neither works.

From these groupings aggregate family earnings and its distribution
for any given year can be constructed provided the following pieces of
information are available for Equation 1.1: (l) the number of marital
labor force participation pairings -- that is, all the possible combina-
tions of husband and wife occupational pairings if they are employed as
well as unemployed or not in the labor force employment statuses; (2)
hourly earnings and hours worked by sex and occupation; and (3) the marital
labor force participation rates -- that is, the number of families who
find themselves in a particular marital labor force participation pairing
as in (1).

Assuming there are n possible occupations for either the husband and/
or his wife to be employed, plus two additional categories for unemployed
and not in the labor force, then there are (n+2)2 possible marital labor
force participation pairings. For any one marital pairing, the annual
earnings of the husband and/or his wife are approximated by the annual
earnings of a male or female in their respective occupations. Summing

t t . . . . .
Eh and Ew in Equation (1.1), then gives the family's total annual earnings,

given a particular marital pairing. With (n+2)2 possible marital pairings,
there are (n+2)2 possible discrete family earnings levels. Multiplying
each family's annual earnings by the marital labor force participation
rate for that marital pairing gives (n+2)2 subtotals of aggregate family
earnings. Summing these subtotals determines aggregate family earnings.
Reranking the subtotals by family earnings levels from lowest to highest
determines the distribution of aggregate family earnings. Algebraically,
m

AEFAMt = 2: ?7Nt(.Et
7'71 1

t
+
h jEw) (1.2)
13

t . .
where AEFAM aggregate family earnings,

(n+2)2, the possible number of marital labor force partici-

m =
pation pairings,
‘I3 = a particular marital labor force participation pairing,
i = the husband's occupation,
j = the wife's occupation
N = the marital labor force participation rate, and
t = year.

The selected studies of Section 2.0 of this chapter on family income size
distribution tended to emphasize the marital labor force participation
rates, EgNt; while those on labor force participation emphasized the rela-
tionship of the husband's earnings, 1E5, to that of his wife, jE:.
Equation (1.2) can be slightly modified in order to deal with the
questions of the relative earnings between husband and wife as suggested
in the literature, the relative earnings between all males and females in
each occupation, the occupational distribution of all males and females,

and how these related factors affect the distribution of aggregate family

earnings. The annual earnings of either the husband or his wife are the

10

products of his/her hourly earnings in his/her occupation and the number
of hours worked annually. Substituting hourly earnings and hours worked
first in Equation (1.1) and then into Equation (1.2) gives the following
results:

Ht t

m
AEFAMt= z -.—.-Nt.(W + ..tw-n), (1.3)
37-13 1w 1 Hh j w J w

U:

where W represents hourly earnings and H represents the average number of
hours worked annually. The annual earnings of husband and wife in a par-
ticular marital labor force participation pairing in relative terms is

found by factoring out the husband's annual earnings.

AEFAM =_Z_3_-i-j-Ntii{Wh°Hh [1+( t)( t)1}. (1.4)

1

U-
y
23"
p
D"

The last two terms of Equation 1.4 give the relative earnings between a
wife and her husband in a particular marital pairing. Multiplying the
terms inside th br k t b ( Wt Wt)( Ht Ht) dd th ' f1 f

e ac e s y i f/i f i f/i f a s e in uence o
the relative wage between males and females in each occupation to the

model.1 Equation (1.4) then becomes:

t H: t iHh Wt jHW Ht
AEFAM = 2,—7N{(i—- it:-)(----)(it.tw-H)[1+(«j——)( t)].} (1.5)
TT'lJ t f1 f
13 iwf in iwh iHh

Equation (1.5) contains six factors which affect aggregate family earnings

. . . . t t . .
and its distribution: (iWh/iwf), the relative hoursly earnings between

males and females in the husband's occupation; (i H h/i H the relative

f)!

number of hours worked annually between males and females in the husband's

. t t . , . . .
occupation; (iwf-in), a floor on any one occupation 5 earnings Since it

 

1Instead of putting the relative hourly earnings and hours worked in
terms of the husband's occupation, the wife's annual earnings could have
been factored out. This would have put the related factors in terms of
the wife's occupation. The results are the same.

11

represents what a female could earn in the husband's occupation;
(jwt/iw:)' the relative hourly earnings between wives and husbands in a

t .
Ht/iHh)' the relative number of hours worked

particular marital pairing; (j w

annually by wives and husbands in a particular marital pairing; and ith'
the marital labor force participation rates. The first three of these
factors are determined by the overall labor force, while the last three
are specific to all husband and wife families. A change in any one of
these factors will affect aggregate family earnings and its distribution.
Whether changes will have an equalizing or disequalizing impact, however,
cannot a priori be determined.

Gini coefficients are used to measure the relative equality or in-
equality of aggregate family earnings. The impact of various factors in
Equations (1.2), (1.3) and (1.5) is determined by holding each factor con-
stant while all else varies, resulting in an adjusted Gini coefficient.
Given a knowledge of the general direction of change in each of these
factors, the impact of any one factor is deduced by comparing the adjusted
Gini coefficient with the actual Gini coefficient. If the former is
greater than the latter, the related factor held constant had an equalizing
impact on the distribution of aggregate family earnings. Using one of the
related factors in Equation 1.5 as an example, if the relative hourly
earnings between all males and females is found to be decreasing and that
the adjusted Gini coefficient with this factor held constant is greater
than the actual Gini coefficient, then it can be surmized that the decrease
had an equalizing impact on the distribution of family earnings -- and

conversely.

12

4.0 The Outline of This Study

Before aggregate family earnings and its distribution can be constructed,
hourly earnings and annual hours worked for all males and females in the
labor force by occupation are developed from Census data. This is done in
a format very similar to Equation (1.1). Within this framework three basic
questions relevant to Equation (1.5) can be answered: (1) What has been
the overall direction of change in the relative hourly earnings between
all males and females in the labor force; (2) What has been the overall
direction of change in the relative number of hours worked between all
males and females; and (3) What has been the overall direction of occupa-
tional shifts for all males and females. Chapter 3 is a summary of these
findings.

Chapter 4 deals with three questions relating specifically to the
population of all husband and wife families. Again following the format
of Equation(1.l)the following questions relevant to Equation (1.5) can be
answered: (1) What has been the overall direction of change in the rela-
tive hourly earnings between wives and husbands and how closely are the
two associated; (2) What has been the overall direction of change in the
relative number of hours worked between wives and husbands; and (3) What
has been the overall direction of occupational shifts for wives as well
as husbands. The findings with respect to these three questions imme-
diately follow the development of the model.

Chapter 5 brings together the information and findings of Chapters 3
and 4 to construct aggregate family earnings and its distribution. An
examination of the various impacts of the six related factors found in
Equations (1.5) is preceded, however, by an examination of the Gini coef-

ficients obtained using Equations (1.2) and (1.3). This is done for two

13

reasons. First, from Equation (1.2) the underlying distributions of all
wives' earnings and all husbands' earnings can be compared. Second, using
quintile analysis the contributions of wives as compared to that of their
husbands in terms of hours worked and earnings received can also be examined.
These extensions provide a more in depth understanding of the Gini coef—
ficients that follow.

Finally, summaries of all the major findings are presented in Chapter
6. The conclusions that will be drawn from these findings will, no doubt,

suggest some policy implications and additional avenues for future research.

CHAPTER 2

RELATIVE HOURLY EARNINGS AND HOURS WORKED
BETWEEN ALL MALES AND FEMALES IN THE LABOR FORCE

1.0 Introduction

 

This Chapter provides an empirical foundation to examine three related
questions. First, what are the recent trends in: (a) the ratio of aggre-
gate female earnings to that of male earnings; (b) the relative hourly
earnings between all females and males; and (c) the relative quantities
of labor supplied by both sexes? Secondly, what proportion of the gap
between aggregate female earnings and male earnings is attributable to
the other trends -- changing relative hourly earnings between females and
males and/or the relative quantities of labor supplied? Finally, in the
event that changing relative hourly earnings and/or relative quantities
of labor are contributing substantially to the earnings gap between fe-
males and males, what proportion of these changes are reflections of occu-

pational shifts in the labor force?

2.0 Aggregate Earnings

 

Aggregate earnings in the economy (AB) is identically equal to the
aggregate annual earnings of all males (EM) and the aggregate annual
earnings of all females (BF). Furthermore, the aggregate annual earnings
of either males or females is equal tothe product of hourly earnings (W)
and the average number of hours worked annually (H); algebraically,

t t t t t
AB 3 W °H + w °H (2.1)
m m f f

14

15

where males are represented by the subscript m; females by the subscript f;
and the time period by the superscript t. Dividing both sides of Equation

(2.1) by the aggregate earnings of all males gives:
t t - t t t t
= + . . o 0
AE /EM 1 [(wf Hf)/(Wm Hm)] (2 2)
Rearranging Equation (2 .2) and solving for Wig/W3:

tt_ tt_ tt
Wf/Wm - [(AE /EM 1)/(Hm/Hf)]. (2.3)

Finally, substituting AEt E EMt + EFt into Equation (2.3):

tt_ tttt
Wf/W - [(EF /EM )/(Hm/Hf)]' (2.4)

Equation(2$9 thus contains all three of the desired trend ratios:
(EFt/EMt), the ratio of aggregate female earnings to those of all males;
(Hi/HE), the relative number of average hours worked per year or the rela-
tive quantities of labor supplied by both sexes; and (WE/Wi), relative
hourly earnings between the sexes.

Annual estimates for the median EFt and EMt are readily available.
Estimates for either Hi/H: or WE/WE, however, are not easily accessible and
must be constructed from disaggregate data. This is done by disaggregating
the male and female labor forces along either occupational or industrial
lines. One delineation serves only as a check upon the other.

Occupational disaggregation involves labeling workers according to
their skill level or occupation. Eleven such categories are used to encom-
pass the 1abor force: (1) Professional, technical, and kindred workers;
(2) Managers, officials, and proprietors, except farm workers; (3) Clerical
and kindred workers; (4) Sales workers; (5) Craftsmen, foremen, and kin-
dred workers; (6) Operatives and kindred workers; (7) Private household

workers; (8) Service workers, except private household workers; (9) Laborers,

16

except farm.and mine; (10) Farmers and farm managers; and (11) Farm laborers
and foremen. Industrial disaggregation divides all workers into various
industries regardless of their occupation. All of the two-digit Standard
Industrial Classifications (S.I.C.) can conveniently serve this purpose.
The classifications are as fellows: (1) Agriculture; (2) Mining; (3) Fores-
tries and fisheries; (4) Construction; (5) Manufacturing; (6) Transporta-
tion, communication, and (ther public utilities; (7) Wholesale and retail
trade; (8) Services and finance; and (9) Public administration.
Disaggregated hourly earnings by sex for these occupations and indus-
tries do not exist. Information on the average number of hours worked per
year by sex, however, can be derived from the data that are available on
the average number of hours worked per week and the average number of weeks
worked per year by sex fbr both occupations and industries. The total
number of hours worked per year by either sex in a particular disaggregated
category is equal to the product of the average number of hours worked per
week and the average number of weeks worked per year by an individual times
the number of males or females employed in that category. Aggregating by
sex over all occupations or industries gives the total number of hours
supplied annually by all males, Hg; or the total number of hours supplied
The ratio of H; to Ht

t
annually by all females, H can be expressed as

 

f' f
follows:
n
t z (.prt - WPY )' Qt
H , m i m
.3 = 1=1
Ht n t t t ’ (2.5)
f .X (iHow ’ WPYf)'iQf
i=1
where H = the total number of hours supplied,

HPW = the average number of hours worked per worker per week,

WPY

the average number of weeks worked per worker per year,

17

Q = the total number of workers employed,

m = males,

f = females,

t = year,

. _ .th .

1 - the l d1saggregate category,

n = 11, for occupational disaggregates, and
n = 9, for industrial disaggregates.

By substituting Equation (2.5) back into Equation (2.4), aggregate relative
hourly earnings between all females and malesiJlthe labor force, Wf/Wi, is
determined. With the appropriate data over some time period, all three
trends in the first question can be examined.

The second question, however, necessitates minor modifications of the
model represented by Equation (2.4). Taking the first difference of the

natural log and rearranging:

t t t t t t
ALn(EF /EM ) - Ammf/wm) - ALn(Hm/Hf). (2.6)

Equation (2.6) gives the percentage change in each one of the trend ratios:
t t t t t t . . . . .
BF /EM , Hm/Hf, and Wf/Wm. By finding the var1ances of each term, it is
possible to determine how much of the variance in ALn(EFt/EMt) or the
aggregate earnings gap between females and males is accounted for by the
variances of the two terms on the right-hand side of Equation (2.6):
Var[ALn(EFt/EMt)] = Var[ALn(W§/W;)] + varLALn(H;/H:)]

tt tt
- A
2Cov[ Ln(Wf/Wm),ALn(Hm/Hf)], (2.7)
then dividing each side of Equation (2.7) by the left-hand term,

tt tt
1 = VarlALn(Wf/Wm)] + Var[ALn(Hm/Hf)]
Var[ALn(EFt/EMt)] Var[ALn(EFt/EMt)]

 

 

t t
-2Cov[ALn(W /W ),ALn(Ht/H
f m m

"hfi'

)1

 

(2.8)
Var[ALn(EFt/EMt)]

18

Equation (2.8) tells how much of the variance of the percentage change

in the ratio of aggregate female earnings to male earnings can be explained
by the variance of the percentage change in the relative hourly earnings
between females and males, by the variance of the percentage change in

the relative quantities of labor supplied by both sexes, and by the covar-
iance between the relative hourly earnings and quantities of labor supplied.
In short, the development of Equation (2.8) should indicate which forces --
relative hourly earnings, relative quantities of labor supplied, or the
interaction of the two -- are having the greatest impact.

Finally, given that relative hourly earnings between females and
males, or their respective quantities of labor supplied in the labor force,
are substantially influencing aggregate earnings, indices are constructed
to measure the direction and magnitude of these changes manifested in occu-
pational shifts within the labor force. Occupational shifts are typically
defined as movements from higher paying occupations to lower paying occu-
pations. The indices constructed to measure these shifts are comparable
to Paasche and Laspeyres indices. The average number of hours worked and
hourly earnings obtained from Equation (2.4) are employed in constructing
these indices rather than normal prices and quantities sold. The indices
are supplemented by another group of Paasche and Laspeyres indices con-
structed with annual earnings and employment figures.

For instance, the Laspeyres index for male hourly earnings is con-
structed in the following manner. First, hourly earnings for the current
periods are multiplied by the base number of hours worked in each industry
or occupation and then summed. This estimate is in turn divided by the

sum of hourly earnings for the base period times the base number of hours

worked in each industry or occupation; algebraically,

19

l-"
II M5
2
Cl:

 

LWINDEX; = , (2.9)

M5
8
:1:

i=11 3 i j
where LWINDEX = the Laspeyres hourly earnings index,
W = the hourly earnings per worker,
H = the average number of hours worked per worker annually,
t = the year,
i = the 1th disaggregate category,
n = 11, for occupational disaggregates,
n = 9, fbr industrial disaggregates,
j = 1, for males,

2, for females.

3'
Similarly, an annual earnings Laspeyres index for males and females is

developed using annual earnings and the actual number of workers employed:

 

LEINDEXE , (2.10)

where LEINDEX = the Laspeyres earnings index,
E = the annual earnings per worker,
Q = the number of workers employed,
t = the year,
i = the ith disaggregate category,
n = 11, for occupational disaggregates,
n = 9, for industrial disaggregates,
j = 1, for males,

j = 2, for females.

20

As a result, there are eight Laspeyres indices: a male hourly earnings
index for industries, a male hourly earnings index for occupations, a male
annual earnings index for industries, a male annual earnings index for
occupations, and a comparable set of indices for all females.

The Paasche indices for hourly earnings are constructed in a similar
manner. First the current period hourly earnings are multiplied by the
current number of hours worked in each industry or occupation and summed.
This estimate is in turn divided by the sum of the base year hourly earnings

times the number of hours currently worked in each occupation or industry;

 

algebraically,
" t t
2 iW3. - 1HJ
PWINDEXE = 1:1 , (2.11)
1 t

where PWINDEX

the Paasche hourly earnings index,
W = the hourly earnings per worker,
H = the average number of hours worked per worker annually,
t = the year,
i = the 1th disaggregate category,
n = 11, for occupational disaggregates,
n = 9, for industrial disaggregates,
j = l, for males,
j = 2, for females.
As before, an annual earnings Paasche index for males and females is

developed using annual earnings and the actual number of workers employed:

.EF - .QF
PEINDEXj = , (2.12)

H'
I) M:

 

IIMZS
m
)0

H.
)-‘
1.)
LJ. H
g.)
l.)

21

where PEINDEX = the Paasche earnings index,

E a the annual earnings per worker,

Q - the number of workers employed,

t a the year,

i = the ith disaggregate category,

n = 11, for occupational disaggregates,

n = 9, for industrial disaggregates,

j = 1, for males,

j = 2, for females.
There are eight Paasche indices as there are with the Laspeyres indices,
making a total of 16 indices for comparison. The indices can be compared
to see whether the Laspeyres indices are greater than the Paasche indices.
If the female Paasche indices are greater than the female Laspeyres indices
then females are moving into relatively higher paying occupations or in-
dustries. If the female Paasche indices are smaller than the female
Laspeyres indices, then females are moving into relatively lower paying
occupations. The same analysis is true for the male indices.

The Paasche and Laspeyres indices show the direction of occupational
or industrial shifts, while the weighted averages that follow show their
impact. The first step in ibis process is to calculate the relative hourly
earnings between females and males solely from disaggregate data. Hourly
earnings are multiplied by the total number of hours worked per year in
each industry or occupation by sex, summed, and then divided by the number
of hours supplied by all males or females in the labor force. A ratio of

the two is derived as follows:

22

 

 

 

n
2: ,wt - at
. 1 f 1 f
1=1
n
t
W: iilin
.13.... n , (2.13)
wm Z .Wt ° Ht
. 1 m i m
1=1
n
z 1";
i=1

where WE/W; = the relative hourly earnings between all females and males,
W = the hourly earnings per worker,
H = the average number of hours worked annually,
m = male,
f = female,
t = the year,
i = the 1th disaggregate category,

n = 11, for occupational disaggregates,

n 9, for industrial disaggregates.

Once Equation (2.13) has been calculated using hourly earnings and
hours worked for current periods, it is recalculated first using current
hourly earnings and base year hours and then using base year hourly earn—
ings and current hours. The difference in estimates of the relative hourly
earnings between females and males using current period data and some com-
bination of base year data shows the impact of occupational shifting. As
an aside, the current year estimates of WE/W; obtained in the manner serve
also as a check upon those found with Equation (2.4).

The same process is used in finding the annual earnings differential

as opposed to the hourly earnings differential between females and males.
Here, as with the Paasche and Laspeyres indices, annual earnings and the

actual number of workers employed are substituted in Equation (2.13).

23

3.0 The Data—Sources, Comparability, and Alterations
3.1 The CPS

Although the data in this study were taken from a variety of published
and unpublished sources, they are comparable. All of the data were directly
or indirectly obtained from the Current Population Survey (CPS) of the
Bureau of the Census and the Department of Labor. The CPS is conducted
each month with approximately 50,000 households representing the entire
population in the United States. The monthly statistics on employment,
unemployment, and related subjects are analyzed and published in such pub-

lications as Employment and Earnings and the Monthly Report on the Labor

 

Force. Once each year in March, information on annual income is also
collected and published in several additional sources. The Bureau of
Labor Statistics (BLS) publishes some of these findings in their Special

Labor Force Reports, while the Census Bureau publishes the Current Popula-

 

tion Reports. The CPS is the only comprehensive survey of this nature.

The income data obtained in household interviews of this type are
subject to various types of reporting errors which tend to produce an
underestimate of income. It is estimated that the income surveys conduc-
ted by the Census Bureau during the past years have obtained about 87
percent of the comparable total money income aggregates and about 95 per-
cent of the comparable money wage or salary income aggregates included
in the personal income series published by the Office of Business Economics
([12], 0.8. Dept. of Commerce, No. 80, p. 10). One such response or re-
porting error stems from the fact that entries are based on memory; in
particular, upon the memory of the wife since she is usually at home during
the time of the interview. Often minor or irregular sources of income are

forgotten or the concept of income is misunderstood by the respondent.

24

Sampling variability in these statistics, however, can virtually be
ignored when major national aggregates are being considered. The chances
are about 68 out of 100 that an estimate from the sample would differ from
a complete census figure by less than the standard error. The chances are
95 out of 100 that the difference would be less than twice the standard
error ([12], U.S. Dept. of Commerce, No. 80, p. 12). Reliability depends
upon both the size of the percentage as well as upon the size of the total
upon which the percentage is based.

Prior to 1967, the sample was spread over fewer areas with fewer inter-
views. Historical data have not been revised to take account of the dif-
ferences because the differences themselves are relatively small. For
most analytical purposes, the data may be regarded as reasonably comparable
([10], U.S. Dept. of Commerce, No. 22, p. 18).

3.2 Sources

The aggregate data necessary for Equation (2.4) are obtained in vari-
ous annual issues of Income of Families and Persons in the United States
[12]. Aggregate annual wage or salary income for all males and females
is approximated by multiplying the median annual wage or salary income
by the number of males and females who had this type of income. The defi—
nition of wage or salary income used by the CPS includes wages, salaries,
commissionships, or pay in kind from a private employer or from a govern-
mental unit. Excluded, however, are wages in kind such as the value of
food and fuel produced and consumed on farms, the net rental value of
owner-occupied homes, the property income received by mutual life insur—
ance companies, and the value of the services of banks and other financial
intermediaries rendered to persons without the assessment of specific

charges.

25

Aggregate annual earnings for all males and females is approximated
by multiplying nonfarm and farm self-employment income by the number of
males and females who had these types of income and then adding them to
aggregate annual wage or salary income. The data for industries exclude
the nonfarm and farm self—employment income; therefore, the data set for
occupations is different from that collected for industries. Given the
importance of wage or salary income as a component of earnings, however,
the two data sets not only serve as a check upon each other, but also the
industrial data should prove insightful with respect to the occupational
data.

In addition, this study uses the only measure available in employing
medians rather than means for income data. The problem can be shown to be
of minor importance in Table 2. If it can be shown that a relatively con-
stant difference exists between the mean and median with respect to wage
or salary income and earnings, then as long as the study confines itself
to relative as opposed to absolute changes, medians are useful. Table 2
shows that this is indeed the case. Presented are both the median and
mean wage or salary income for all industries and occupations as well as
the comparable earnings data. Taking the ratio of mean to median and then
wage or salary income to earnings income from 1967 to 1971, gives a con-
stant relationship between the two. The final average of 1.0281 for males
and 1.0139 for females reveals that the scalar difference is small. This
fact enables the assumption to be made that the bias inherent in using
medians rather than means is irrelevant for the purposes of this study.

The disaggregate median annual wage or salary income by sex and industry
and median annual earnings by sex and occupation were also obtained from

Income of Families and Persons in the United States [ll], [12]. It is

26

TABLE 2

A Comparison of Median and Mean Incomes

Wage or Salary Income--A11 Industries and Occupations [7]

1967
1968
1969
1970
1971

1967
1968
1969
1970
1971

 

 

Male Female
Median Mean Median Mean
$5,974 $6,116 $2,295 $2,678
6,376 6,523 2,468 2,869
6,854 7,105 2,612 3,089
7,108 7,469 2,801 3,341
7,351 7,774 3,045 3,541

Earnings-~A11 Industries and Occupations

 

Male Female
Median Mean Median Mean
$6,020 $6,337 $2,351 $2,779
6,442 6,811 2,512 2,921
6,899 7,340 2,564 3,091
7,152 7,685 2,730 3,328
7,388 8,023 2,986 3,524

Ratio of Mean to Median Incomes--A11 Industries and ccupations

1967
1968
1969
1970
1971

 

 

Wage or Salary Income Earnings
Male Female Male Female
1.0237 1.1668 1.0559 1.1820
1.0230 1.1624 1.0572 1.1628
1.0366 1.1826 1.0639 1.2055
1.0508 1.1928 1.0745 1.2190
1.0575 1.1629 1.0860 1.1802

Ratio of Mean/Median Wage or Salary Income to
Mean/Median Earnings-~All Industries and Occupations

Male Female
1967 1.0314 1.0130
1968 1.0334 1.0003
1969 1.0263 1.0193
1970 1.0225 1.0219
1971 1.0269 1.0148
AVG. 1.0281 1.0139

27

of interest at this point to note that the statistics that will be used

for comparison in Chapter 3 from the U.S. Department of Labor/Workplace

Standards Administration/Women's Bureau are based upon the same set of data.
The data used in calculating the relative number of hours supplied by

males and females, Hi/HE, were found in two separate sources. First,

Employment and Earnings [16] provided the average number of hours worked

 

per week per person by sex and occupation for the years 1967 through 1971.
BLS provided the unpublished data for the years 1962 through 1966. BLS also
provided the average number of hours worked per week per person by sex for
all industries from 1962 to 1971 [12]. The estimates for the average num-
ber of hours worked per week per person may differ somewhat from other
sources since the estimates are based on the actual number of hours worked
during the survey week. For persons working more than one job, the figure
relates to the number of hours worked at all jobs during the E§§_survey
week, and all the hours are credited to the longest job.

The second source of information is The Werk Experience of the Popula-
Eigg_[l9]. The average number of weeks worked per year per person by sex
for industries or occupations is obtained from various Special Labor Force
Repgrts. Published data cover the period of 1962 through 1970. BLS again
furnished the unpublished data for 1971. The published form of the average
number of weeks worked per year per person is disaggregated into two groups:
full-time and part-time. Full—time workers are those who work 35 or more
hours per week while those classified as part-time work less than 35 hours
per week. Each one of these groups is further divided into three more
groups: those who worked 50-52 weeks per year, those who worked 26-49
weeks per year, and those who worked 1-26 weeks per year. Thus this source

not only provided the number of males and females employed in a particular

28

industry or occupation, but also their status with respect to the number
of weeks worked per year per person.
3.3 Alterations

Slight alterations in all four series -— wage or salary income data
and earnings data, hours worked per week per person data, weeks worked per
year per person data, and the number of workers employed data -- are neces-
sary. In order to obtain a complete and consistent data set, the first
compromise involved the average number of hours worked per week per person.
The unpublished industrial data received from BLS combined both Agriculture
and Mining in the early years, 1962 through 1966. The constraints imposed
upon the other three industrial data series were complicated further by

the median annual wage or salary income estimates found in the Current Popu-

 

lation Reports. This data series combined Agriculture with Forestries and

 

Fisheries. Given the significance of Agriculture in terms of the number
of people employed, all three industries were combined leaving only seven
industrial categories. The extent of the disaggregating problem is easily
shown. Over the 10-year period of 1962 through 1971, Forestries and Fish-
eries was incorporated in three different S.I.C.'s. On several occasions,
Forestries and Fisheries is found in Service and Finance, with Agriculture,
or by itself. As a result all four data series for industries -- median
annual wage or salary income, hours worked per week per person, weeks
worked per year per person, and the number of workers employed -- were
adjusted by disaggregating and reaggregating using employment figures as
weights to make Agriculture, Forestries and Fisheries, and Mining one
classification.

Occupational data was less troublesome. Private Household workers

were sometimes included in the Service occupation. After excluding them

29

from Service to form their own category, the earnings, hours, and employ-
ment occupational data sets were compatible.

After the above alterations were made, the average number of weeks
worked per year per person was tabulated by sex and industry or occupation
from the disaggregate data. The average number of weeks worked per year

per person is found by taking a weighted average as follows:
.WPY? = (l-.YF)(51-.A1F+38-.A2F+13-.A3F)+.YF(51-.Bli+38-.82?
1 J 1 J 1 J l J 1 J l 3 1 J 1 J
t
+13-.B3.), (2.14)
l 3

where WPY = the average number of weeks worked per year per person,
y = the percentage of part-time workers,

A1 = the percentage of full-time workers who worked 50-52 weeks
per year,

A2 = the percentage of full-time workers who worked 27-49 weeks
per year,

A3 = the percentage of full-time workers who worked 1-26 weeks
per year,

B1 = the percentage of part-time workers who worked 50-52 weeks
per year,

82 = the percentage of part-time workers who worked 27-49 weeks
per year,

B3 = the percentage of part-time workers who worked 1-26 weeks
per year,

i = the ith disaggregated category,
n = 11, for occupational disaggregates,
n = 7, for industrial disaggregates,
j = 1, for males,
j = 2, for females,
t = the year.
The midpoint of the number of weeks worked per year, i.e., 51 weeks for

those full- and part-time workers who worked 50-52 weeks per year is chosen

30

as the appropriate weight. Equation (2.14) is exact for all seven of the
industrial disaggregates but is modified for occupational disaggregates.
The work experience weekly breakdown for part-time workers did not exist
for the entire 10-year period. Thus it was assumed that all part-time
workers fall within the 27-49 weeks worked per year category. This is not

an unlikely assumption, since few workers are found in the extremes;

algebraically,
.wpy? = (1-.y§)(51-.A1§+38-.A2?+13-.A3F) + ,yF-3s. (2.15)
1 j 1 J 1 J l J l J 1 J

4.0 Summary

At the outset of this chapter, three interrelated questions were
posed. This chapter not only proposed a method of analysis, but also the
data to be used. The next chapter will present the findings and an analysis
as to how they relate to each other. Of greater importance is the fact
that these findings will in turn provide the foundation upon which to
determine the possible impacts of working wives and related factors upon
the distribution of family earnings for all husband and wife families.
The relative hourly earnings, the relative number of hours supplied annu-
ally by both sexes to the economy, as well as their trends, will play an

important role in achieving this objective.

CHAPTER 3

AGGREGATE TRENDS AND THE IMPORTANCE OF
RELATIVE CHANGES IN THE LABOR FORCE

1.0 Introduction

The preceding chapter posed three related questions: (1) What are
the recent trends in the ratio of aggregate female earnings to male earnings,
the relative hourly earnings between females and males, and the relative
number of hours supplied by males and females; (2) What proportion of the
gap between aggregate female and male earnings is attributable to the
trends mentioned in (1); and finally, (3) What proportion of these trends
are reflections of occupational shifts within the labor force. This chap-
ter will present and discuss in detail the empirical findings. For pur-
poses of comparison and discussion, these findings will be supplemented
with information contained in the Handbook of WOmen Workers [20]. In addi-
tion to the information contained in this Women's Bureau publication, their
data were updated in order to make a direct comparison over the 1962 to

1971 period possible.

2.0 Aggregate Trends in Earnings, Hours, and Wages
2.1 Earnings and Wage or Salary Income

The ratio of aggregate female earnings to that of all males in the
labor force increased noticeably over the period of 1962 to 1971. The
annual estimates for EFt/EMt found for Equation (2.4) using occupational

data rose from .20 in 1962 to .26 in 1971. Similarly, the ratio of aggregate

31

32

female wage or salary income to that of all males also rose from .20 in

1962 to .27 in 1971 as shown in Table 3. The estimates indicate that

TABLE 3

The Ratio of Aggregate Female Earnings to That of Males
and the Ratio of Aggregate Wage or Salary Income
of Females to That of Males, EFt/EMt, 1962 to 1971

 

Year Earnings Wage or Salary
1962 .20 .20
1963 .19 .21
1964 .20 .22
1965 .22 .24
1966 .20 .23
1967 .22 .25
1968 .23 .25
1969 .24 .25
1970 .25 .26
1971 .26 .27

aggregate earnings and wage or salary income of all females is growing at

a proportionately faster rate than the aggregate earnings and wage or salary
income of all males in the labor force. This suggests the following five
possibilities:

l) The relative quantity of labor supplied by males to that supplied
by females is decreasing, ceteris paribus;

2) The relative hourly earnings or wage rate between females and
males is increasing, ceteris paribus;

3) The relative quantity of labor supplied by males to that supplied
by females is decreasing, while the relative hourly earnings or
wage rate between females and males is increasing but at either
a proportionately faster or slower rate;

4) The relative quantity of labor supplied by males to that supplied
by females is decreasing, while the relative hourly earnings or
wage rate between females and males is decreasing, but at a propor-

tionately slower rate; and

33

5) The relative quantity of labor supplied by males to that supplied
by females is increasing, while the relative hourly earnings or
wage rate between females and males is increasing at a proportion-
ately faster rate.

Any other combination of increasing or decreasing relative factors would
lead to a downward drift in EFt/EMt. An examination of the relative quan-
tity of labor supplied by males to that supplied by females will eliminate
some of these possibilities.
2.2 The Relative Quantity of Labor Supplied

Aggregating the number of hours worked per week, the number of weeks
worked per year, and the number of male or female workers employed across
occupational and industrial lines shows that the relative quantity of annual
hours supplied by all males to that supplied by females is decreasing.
These findings immediately eliminate possibilities (2) and (5). Table 4

. . . . t t
presents the occupational and industrial est1mates for Hm/Hf.

TABLE 4

The Relative Number of Annual Hours Supplied by All Males
to That Supplied by All Females From Occupational
and Industrial Categories, Hfi/HE, 1962 to 1971

 

 

 

Year All Occupations All Industries
1962 2.47 2.46
1963 2.47 2.42
1964 2.44 2.48
1965 2.42 2.41
1966 2.35 2.35
1967 2.30 2.27
1968 2.28 2.26
1969 2.23 2.22
1970 2.20 2.21
1971 2.19 2.17

Males in general supply at least twice as many hours as do females to

the economy each year, but the proportion is continually decreasing.

This

34

implies a disproportionate increase in the relative number of hours supplied
by females. Indeed, the literature on labor force participation rates for
women would imply such a trend. The discrepencies between the columns for
all occupations and all industries is a result of the two different data
sets. The exclusion of those workers with farm and nonfarm self-employment
income from the industrial data provides an additional comparison. In gen-

. . . t t
eral the 1ndustr1al estimates of Hm/H

f are lower than those for occupations,

suggesting that females are working proportionately more hours for wage or
salary income than for self-employment income -- and conversely for males.

The literature, however, would also imply that the decreasing trend
in the relative number of hours supplied by all males to that supplied by
all females is a result of the increased labor force participation of fe-
males; and in particular, to married females. No doubt the increased rela-
tive employment of females to males has played a substantial part in in-
creasing the number of hours supplied by all females, but changes in the
relative number of hours worked per week, the relative number of weeks
worked per year, and thus the relative number of hours worked per year
between males and females may be of equal significance.
2.2.a The_Quantity of Labor Supplied by Occupations

Over the 10-year period of 1962 through 1971 for all occupations, as
summarized in Table 5, the average percentage change in the number of hours
worked per week per person was negative for both sexes. The overall occu-
pational average decrease for males was only .18 percent. On the other
hand, the average percentage decrease in the number of hours worked per
week per person for females was .50 percent. Table A in Appendix A indi—
cates that females in general worked fewer hours per week per person than

males. Table G exemplifies the point in relative terms. As a result the

average decrease in the number of hours worked per week per person for

35

TABLE 5

Average Annual Percentage Changes in Selected
Variables by Occupation, 1962 to 1971

 

  
   
 

 

 

 

 

 

 

 

 

 

 

 

 

l o
Variable 44 m. m -§ é.$
0:. 03. 03. 2: 2“
H m o $404C) H m o n. g g
o m o m o m In .4 c
.g'o H .g'o w .g'o H 0'0 6 m -a
m m o o w w m u n c
.x m .x m .x m H >. o s H c
z u 2 H 2 H m o B o m o
o u o H o H .4 m u m m
m 3 m m 3 w m 3 o 'g o. m m N Sex
6) m c) m c) m 5 gnu o >4m
m m m m m m z m 0:» .4 m
Occu- H ;.x H.¥ H H u w u u H
4.8 :33 233 323 4““ 5“
patlon «:23 43>. <32» am» «2:33 £3
Professional -.37 -.18 -.55 3.25 2.69 6.11 Male
-.35 -.27 -.09 4.55 4.46 6.14 Female
Mana ers -.44 -.04 -.48 .23 -.27 7.31 Male
9 -.77 .12 -.64 3.43 2.68 8.60 Female
Clerical -.39 -.56 -.95 1.01 .04 5.55 Male
-.48 -.25 -.72 4.12 3.36 4.72 Female
S les .39 1.08 1.48 1.43 2.84 4.01 Male
3 -.84 -.O6 -.89 1.36 .47 4.29 Female
Craftsmen -.13 .08 -.05 2.65 2.60 4.91 Male
-.55 -.37 -.99 6.27 4.71 5.06 Female
0 ratives -.12 -.11 -.23 .55 .35 4.66 Male
pa -.09 .20 .11 1.74 1.88 4.79 Female
Private 1.26 -- -- 1.87 -- -— Male
Households -.34 .03 -.31 -4.96 -5.18 3.45 Female
Service -.71 -.81 -1.52 4.15 2.58 3.92 Male
-1.18 .02 -1.16 4.62 3.41 6.40 Female
Laborers .12 .22 .33 1.74 2.01 3.08 Male
-l.04 .46 -.55 18.74 16.56 9.82 Female
F ers .27 -.15 .12 -3.64 -3.51 6.54 Male
.66 .51 1.14 5.06 6.51 6.13 Female
Laborers .55 -.14 .43 -4.92 -4.53 2.29 Female
*All Across -.18 —.03 -.22 1.39 1.16 5.17 Male
Time -.50 -.01 -.52 2.82 2.27 5.15 Female

 

 

 

*This figure was derived by holding the
weighted average for all occupations.

year constant and then taking a

36

females is greater in absolute terms, but may not be greater in relative
terms. Nonetheless, the overall average decreases are due to the negative
average percentage change in the number of hours worked per week per person
in six out of the 11 occupations for males, and nine out of the 11 occupa-

tional categories for females.

Looking at specific occupations, the largest average decrease in the
number of hours worked per week per person for males was in Service workers,
except Private Household workers with a .71 percent decrease. In contrast,
one of the largest average increases for males occurred in Sales workers
with .39 percent. Although Private Household workers also had an average
increase of 1.26 percent, a much smaller proportion of the male labor force
is situated in this occupation. The Sales workers' occupation is worthy
of further note since females registered one of their larger average per-
centage decreases here. The .84 percent average decrease was sizeable,
but Service workers, except Private Household workers and Managers, Offi-
cials, and Proprietors, except Farm Wbrkers, also recorded average decreases
of 1.18 and .77 percent respectively. The only average increases for fe-
males took place in Farmers and Farm Managers with .66 percent and in Farm
Laborers and Foremen with .55 percent. Both are occupations employing a
very small segment of the female labor force. It is safe to say then that
the average decreases for females were approximately equal to and in most
cases greater in absolute terms than those for males. The average decreases
were more widespread than for males resulting in total in a sharper average
percentage decrease.

Tables B and H in Appendix A provide the background data in analyzing
the average percentage change in the number of weeks worked per year per

[Kirson found in the second column of Table 5. The overall average percen-

37

tage change for males and females demonstrates the same behavior. While
males showed a negligible average decrease of .03 percent in the number of
weeks worked per year per person from 1962 to 1971, females showed a
smaller average decrease of .01 percent. Males recorded an average de-
crease in the number weeks worked per year per person in six of the 11
occupations. Females, in contrast, recorded an average increase in six
of the 11 occupations.

In Table 5 the greatest average decrease for males occurred again in
Service workers, except Private Household workers with a .81 percent decrease.
Clerical and Kindred workers also had a sizeable average decrease of .56
percent. The most significant average increase for males, however, was in
Sales workers with 1.08 percent. In this same occupation, females had an
average decrease of .06 percent. Other average decreases for females
occurred in Clerical and Kindred workers with .25 percent, and in Crafts-
men, Foremen, and Kindred workers with .37 percent. The largest average
increases for females were in Laborers, except Farm and Mine with .46 per-
cent and in Farmers and Farm Managers with .51 percent. The percentage
of the female labor force in these last two occupations was a little over
3 percent in 1971.

Given the overall average decreases in both the number of hours worked
per week and the number of weeks worked per year per person for males, it
is not surprising that the average percentage change in the number of hours
worked per year per person was slightly negative. Table C in Appendix A
has the raw data for the estimates found in the third column of Table 5.
Similarly, the average percentage change in the number of hours worked
Per year per person for females was negative resulting from average de-
crteases in seven of the 11 occupations. As can be expected from the hours

{xxr'week data, the average decrease for females is greater than that for males.

38

Finally, in column four of Table 5, the total number of males employed
by all occupations over the decade increased significantly with an average
increase of 1.39 percent. The average percentage increase, however, was
enough to cause the average percentage change in the number of hours sup-
plied by all males in the economy, Hi, to be positive -- 1.16 percent --
despite the average decrease in the number of hours worked per year per
person. The increased number of females in the market allowed for an over-
all average increase of 2.82 percent, which in turn offset the average de-
crease in the number of hours worked per year, resulting in a substantial
average increase in the number of hours supplied by all females in the
economy, HE, of 2.27 percent.

Although the increased labor force participation of females seems to
be the overriding factor in determining the number of hours supplied by
all females, a changing work pattern is also perceivable. A changing
work pattern, hours worked annually, seems to dampen the hours of females
more than males, although in Operatives and Kindred workers and in Farmers
and Farm Managers, the changing work pattern accentuated the increased
labor force participation of females. In Private Household workers, the
decreasing working pattern in addition tothe average annual decrease in
the number of females employed exaggerated the average annual decrease in
the number of hours supplied by all females.

In summary, the component parts of the number of hours supplied by all
males or females reveal definite trends. Both sexes worked fewer hours
per year per person with females working proportionately fewer. The average
increase in the male labor force is not always sufficient to positively
(Iffset this trend while the average increase, at least for this time period,

for females is .

39

2.2.b The Relative Quantity of Labor Supplied by_Occupations

The significance of these changes are further explored by looking at
the average percentage change in the relative number of hours worked per
week per person between females and males, the relative number of weeks
worked per year per person between females and males, the relative number
of hours worked per year per person between females and males, etc. The
underlying tables for all of the variables found in Table 6 are Tables G
through K in Appendix A.

The overall average percentage change for all occupations with respect
to the relative number of hours worked per week per person between females
and males was negative. The average decrease of .14 percent over the 10-
year period verifies the earlier hypothesis that the number of hours worked
per week per person is decreasing faster for females than for males. In
six of the 11 occupations, the average percentage change was negative.
Positive average percentage changes were the greatest in the farm related
occupations.

The overall average percentage changes in the relative number of weeks
worked per year per person, however, was positive; resulting from positive
average increases in seven of the 11 occupations. The most notable average
increase was .86 percent in Service workers, except Private Household. In
contrast, Sales workers showed an average decrease of 1.11 percent. Again
the findings are consistent with the previous section. The effect is a
slight average percentage increase in the relative number of hours worked
per year per person of .02 percent. This means that on average the number
of hours supplied to the market annually by females is decreasing but at
a relatively slower rate than that of males.

In conjunction with the fact that the average percentage change in

the relative number of females to males employed is positive, at an average

40

TABLE 6

Average Annual Percentage Changes in Selected

Relative Variables By Occupation, 1962 to 1971

 

 

o c
Relative §§ 'Eé 'gg H "
Variable 4“: 58:: 533 4: 53 ..
ZED: 230-: 2304 23% 2mg ‘0;
o H w m o m-a
3‘”: 3:23 egg 2.... .3. >5
-H g m -a m m -H m 'H H m -H -H
p o . u m .u o u g u o u u m
o m.x m 3 H m m H m 'o m m 0 win
.4 0 '4 M '*,H 8 '3.“ g, 'YIH m '3 >1
CC 484 3‘34 :2... .203 80.8 .43
cu-
m H H o H H o H u m H a. o H-H o 5
' .CQJr-i O
patio“ 15.83 15.33 1:533.) 5.85 Econ. ESE:
Professional .02 .44 .47 1.30 1.76 .14
Managers -.34 .19 -.14 3.18 3.04 1.21
Clerical -.07 .29 .23 3.39 3.60 -.89
Sales -1.17 -1.10 -2.24 .20 -2.15 .57
Craftsmen -.44 -.21 -.69 2.68 1.48 .49
Operatives .06 .37 .43 1.24 1.65 .05
Private
Household 1'02 - -- 1°90 -- --
Service -.43 .86 .41 1.16 1.58 2.69
Laborers -.15 .43 .38 10.06 9.67 3.36
Farmers .54 .68 1.20 6.52 8.09 9.81
Farm Laborers .58 -.46 .15 -l.l7 -.79 -3.03
*
All Across -.14 .19 .02 2.40 2.27 .71

Time

 

 

*This figure was derived by holding the year constant for occupations and

then taking a weighted average for all occupations.

41

annual increase of 2.40 percent, the relative number of hours supplied by
all females and males had a 2.2 percent average increase and thus Hi/H:
decreased from 1962 to 1971. Keeping in mind the changing work patterns,
the relative number of hours supplied by all males and females is decreasing
slower than it would if the female working patterns were keeping pace with
those of males. The changing work patterns are plausibly a direct result

of the fact that the increased labor force participation rates of females
are due chiefly to the increased labor ferce participation of married fe-
males. These females have more donfining responsibilities at home allowing
for only part-time work, or full-time with little or no overtime. At the
same time, the average percentage change in the relative number of weeks
worked per year by females and males is positive. A tendency of more weeks
worked per year on the part of females and fewer for males seems to be an
adjustment for some as of yet unidentified change in life styles.

Several of the occupations definitely reflect such changes. The
overall positive average increases in relative variables in Professional,
Technical, and Kindred workers show that more females are willing to put
in longer hours each week and more weeks of the year. A similar situation
exists in Operatives and Kindred workers, Farmers and Farm Managers, Sales
workers, and Craftsmen, Foremen, and Kindred workers which show an adjust-
ment process with a decreasing work pattern in terms of hours across the
occupations. Surprisingly enough, Clerical and Kindred workers have made
very little adjustment.

The disaggregate data show that the average increase in the relative
number of hours worked per year by all females and males is not just a
Phenomena solely related to the increased labor force participation of

women, but is also a result of males working fewer hours per week, fewer

42

weeks per year, while relatively more females are working relatively fewer
hours more weeks of the year. If this is indeed a trend, the industrial
data should augment these findings.

2.2.c The£Quantity of Labor Supplied by Industries

 

Table 7 presents a summary of the industrial data for 1962 to 1971.
The actual data are found in Tables N through S in Appendix A. Column 1
of Table 7 shows that the average percentage change in the number of hours
worked per year per person for both sexes was negative. The overall indus-
trial average decrease for males was only .02 percent. Although females
also had a negative average percentage change in the number of hours worked
per week per person, it was not as great as the occupational data.

The average decrease for males in the number of hours worked per week
per person was a result of average decreases in four of the seven industries.
The largest average decreases occurred in Wholesale and Retail Trade and
in Public Administration with .26 percent and .23 percent respectively.
Manufacturing along with Transportation, Communication, and other Public
Utilities both recorded positive average changes in the number of hours
worked per week per person for males. In these two industries females
had a negative average percentage change. In the two categories with
positive average percentage changes for females, Agriculture, Forestries
and Fisheries, and Mining had the largest with .49 percent. It can be seen
from Table 7 that the average decrease incurred by females was again abso-
lutely greater than that for males, confirming the occupational findings.

The average percentage changes in the number of weeks worked per year
per person, however, were inconsistent with the occupational data. Rather
than a decrease in the average number of weeks worked per year, both males

and females recorded a significant average annual increase of .22 and .63

43

TABLE 7

Average Annual Percentage Changes in Selected
Variables by Industry, 1962 to 1971

 

     

 

 

 

 

 

 

 

 

l 'O
m: m 94 H E.3
o H o H o u o H H
m c m c m c 3 w z o
H 94 O H 94 O H 0-) O Du 3
m m o m m m In .4 o
'O H 'O H 'U H O 'O (U 01 JJ
313 334 321 .9 “a a.
2 H 2 M 2 H d) O (9 O O
.9: .9: .93 g3 .84 32
m a. 0n n. 0) n. E 619-) a) m a) sex
m m m m m m z m 0 >4 3 m
0 __ H 9.54 H .34 H H g H H H
°w 40$ :83 :03 was 433 23
Patlon #3313 fiSW 5‘th 150:» 63.004 904
g .49 3.37 3.90 -3.84 -.66 6.77 Female
Construction -.05 .48 .44 2.91 3.33 5.44 Male
-.20 4.47 4.49 6.26 11.42 .17 Female
. .10 .19 .30 1.08 1.42 4.57 Male
"anufaCturlng -.06 .14 .08 1.98 2.07 4.80 Female
Public .02 -.13 -.12 2.12 1.97 5.94 Male
Utilities -.58 .26 -.31 4.06 3.70 5.53 Female
-.86 .20 -.66 3.25 2.58 4.48 Female
Service .05 .46 .52 3.38 3.89 5.43 Male
.25 .95 1.22 3.55 4.78 6.14 Female
Public -.23 -.01 -.24 2.03 1.75 6.98 Male
Administration -.21 -,15 -.33 2.71 2.38 5.66 Female
*All Across -.02 .22 .21 2.18 1.96 5.27 Male
Time -.11 .63 .53 3.59 3.06 5.32 Female

 

 

 

*This figure was derived by holding the year constant for industries and
then taking a weighted average for all industires.

44

percent respectively for those workers who received wage or salary income.
The average percentage change was negative in only three of the seven in-
dustries for males and one of the seven for females.

Behavior is different for those with only wage or salary income as
opposed to those with several sources of earnings again in that the average
percentage change in the number of hours worked per year is positive for
both sexes as found in column 3 of Table 7. The positive percentage changes
in the number of weeks worked dominated the negative percentage changes in
hours worked per week. Average decreases in the number of hours worked
per year occurred in three of the seven categories for males as well as
for females in the same three industries. The greatest average decrease
fOr both males and females was in Wholesale and Retail Trade with a .62
percent decrease for the former and a .66 percent decrease for the latter.
The largest average increases for females was in Agriculture, Forestries
and Fisheries, and Mining as well as in Construction. While Construction
also had one of the largest average increases for males of .44 percent,
Services and Finance had an equal average increase of .52. Agriculture:
Forestries and Fisheries with its 1.39 percent average increase for males
was the largest. Females had sizeable gains in this category and in
Construction.

All but one of the industrial categories had a positive average per-
centage change in the number of workers employed for both sexes. Column 4
of Table 7 shows that the largest average increase for males was in Service
and Finance with 3.30 percent. Substantial average increases also were
recorded in Construction, Wholesale and Retail Trade, and in Public Admin-
istration with 2.91, 2.70, and 2.02 percent respectively for males. Two

different industries besides Construction -- Service and Finance and

45

Transportation, Communication, and other Public Utilities -- showed sizeable
positive average increases for females of 6.26, 3.55, and 4.06 percent re-
spectively. Agriculture, Forestries and Fisheries had the only average
percentage decreases for both sexes.

As a result of these trends, the average percentage change in the
number of hours worked per year by all females and males was positive --
1.96 percent fer males and 3.06 percent for females. Both estimates of
the average increases are higher than the occupational data since a dampen-
ing effect was not felt through a decreasing work pattern. Females had an
average increase substantially above that for males with respect to the
number of hours supplied by them. In short, the number of hours supplied
by all females is increasing at a faster rate than that supplied by males
in the economy, thus Hi/H: is decreasing.

2.2.6 The Relative Quantity of Labor Supplied by Industry

Although the average percentage changes were different in absolute
terms, the same pattern develops in relative terms for industries as in
occupations. The estimates found in Table 8 are based on the raw data
found in Tables T through V in Appendix A. As in the occupational data,
the average percentage change in the relative number of hours worked per
week per person between females and males was negative. This phenomenon
was a result of a relative average percentage decrease for this variable
in four out of the seven industries. Similarly, the average percentage
change in the relative number of weeks worked per year per person was posi-
tive. Relative average percentage decreases occurred in Manufacturing and
in Public Administration. The average percentage change in the relative
number of hours worked per year per person was also positive. This time

Transportation, Communication, and Other Public Utilities in addition to

46

TABLE 8

Average Annual Percentage Changes in Selected

Relative Variables By Industry, 1962 to 1971

 

  

 

     

 

comuwm Hem mama m... B w W mu W Q. U
wmms m>wumHmm 059 .L .L . . a . .L a
. .
Ham.» Hem 60.32 3 2 4 O 5 4 3 5
Im5m $30: no Hon 6. cm 7. n” 9. on 6. 9”
I552 m>wum~om was 6 l 1
use.» Mom cmmoamfim 4 9 3 6 8 5 6 6
23.8.. to .3 a s on a” s a" 7. 7.
I552 w>flumHmmaonB a. 2 1
common Mom .5602 50m 3 2 8 0 l 2 6 7
09.503 95.52 mo Hun 9" a... I... I” J 7. nw A
I852 m>fiumamm one 2 .3 . . .
comuwm 502 5002 Mom 7. I. 2 8 7 0 O 2
6 5 O 4 5 5 l 5
Mumxuoz mxwmz mo H03 L in _o o o o .o o
I852 0>wumamm 02B
sownmm Men 3003 50m 5 S 5 8 S l 2 6
@9303 mH502 mo web a” 7" J a” 9" 9” O. 0.
I552 o>aumamm one . . . _ .
n
e g 0
l n m .l
e e 0 . t S
V a r .1 M a S
.1 .1 y u t S r O
t r r t C t e t r
a a t 1 u C .1 e S C
l V s u r a c t c c.i A
u c .t .r .1.1 e .i .1 n
m .d i. s u .1.1 .d v .1.1 .1 e
n r n .m.1 a r .m.m .1 m
T. a. o t r e A .
A C P U T S P A .a T

 

47

Manufacturing and Public Administration recorded the relative average per-
centage decreases. The relative number of female and male workers employed
was positive except in Agriculture, Forestries and Fisheries, and Mining.
As a result the average percentage change in the number of hours supplied

annually by all females and males was positive, thus Hi/Ht is also decreasing

f
as in occupational data.
2.3 HourlyAEarnings and Wage Rates

Since the overall average percentage change in the relative number of
hours supplied by all females and males is definitely positive, possibili-
ties (l), (3) and (4) still exist. By looking at the trends in the relative
annual hourly earnings and wage rates, for occupations and industries re-
spectively, the possibilities can be narrowed even further. This is done
by solving Equation (2.4) for WE/Wfi. Theoretically, the substitution of
the ratio of aggregate annual female earnings or wage or salary income to
that of all males and the ratio of the aggregate number of hours supplied
annually by males to that of females into Equation (2.4) should give us a
good approximation of the relative hourly earnings for occupations and the
relative wage rate for industries. Unfortunately, the data were not that
accurate. The appropriate approximation of the ratio of aggregate annual
female earnings to that of all males was found in the following manner.

First, by using the estimates of EFt/EMt found in Table 3 and those
for Hi/H: found in Table 5, WE/W; was calculated. Table 9 contains the
results. Hourly earnings seem to exhibit a more noticable upward trend than
the wage rate, although both behave eratically. Comparing these estimates

of hourly remunerations to those calculated in Tables L and Y in Appendix A

. t .
suggests that these estimates of WE/wm are too low. The reason 18 that

t

f observations were dropped that had insufficient

. . t
in the calculation of Hm/H

48

TABLE 9

The Ratio of Hourly Annual Earnings or Wage Rates
Between Females and Males, WE/Wfi,
Using Aggregate Persons' Data, 1962 to 1971

 

 

Year Hourly Earnings Wage Rates
1962 .50 .55
1963 .52 .55
1964 .54 .59
1965 .57 .62
1966 .54 .58
1967 .57 .60
1968 .57 .61
1969 .56 .59
1970 .57 .61
1971 .59 .62

data, thus the aggregate number of males and females with earnings was six
percent larger than the total in the disaggregate categories. The estimates
t t . . .
of EF /EM were too small Since a disproportionate percentage of the dropped
data was in the male labor force. This would bias the estimates of Equation
. . t .
(2.4) downward. A new estimation of EF /EMt was then calculated by multi-
plying annual earnings or wage or salary income by occupation and sex, or
by industry and sex times a weighted average of the number of workers in
. . . t t
that particular disaggregate. The new estimate for both EF /EM and
t t . . t t . .
Wf/W are found in Table 10. The estimates of Wf/W are now more in line
m m
with the underlying raw data and reveal little if any closing of the remun-
erations gap. An examination of particular occupations and industries
t

. . . t
shows the sources of the eratic behaVior in Wf/Wm.

2.3.a Hourly Earnings Rates by Occupation

 

Over the 10-year period the average percentage change in the hourly
earnings for males was positive with an overall occupational average in-
crease of 5.17 percent. Females, despite their lower initial wage, had a
5.15 percent average annual increase. The largest average increase for

males was in Managers, Officials, and Proprietors with a 7.31 percent

49

TABLE 10

The Ratio of Aggregate Female Earnings to That of Males
and the Ratio of Aggregate Wage or Sala Income
of Females to That of Males, EFt/EM ,
Using Weighted Averages, 1962 to 1971

 

Annual Annual Wage or
Year Earnings Salary Income
1962 .26 .30
1963 .27 .30
1964 .25 .31
1965 .29 .32
1966 .29 .32
1967 .30 .33
1968 .30 .34
1969 .30 .34
1970 .31 .34
1971 .31 .35

The Ratio of Annual Hourly Earnings or Wage Rates
Between Females and Males, WE/Wg,
Using Weighted Averages, 1962 to 1971

 

Year Hourly Earnings Wage Rates
1962 .66 .75
1963 .66 .73
1964 .67 .76
1965 .70 .77
1966 .67 .76
1967 .68 .76
1968 .69 .77
1969 .66 .75
1970 .67 .75

1971 .68 .76

50
average increase. The next highest were recorded in Farmers and Farm Mana-
gers with a 6.53 percent average increase and Farm Labor with a 6.66 percent

average increase. The other average annual increases for males ranged

from 3 to 6 percent. Females had their largest percentage increase in
Nonfarm Labor with a 9.82 percent increase as shown in column 6 of Table 4.
This average increase is 6 percent greater than that of their male counter-
parts. A similar situation occurred in Service workers, except Private
Household workers. The opposite phenomena was true in Farm Labor. In
Clerical and Kindred workers where a substantial portion of the female
labor ferce exists a smaller average increase of 4.72 percent was recorded
as compared to that of males -- 5.55 percent.
2.3.b Relative Hourly Earnings By Occupation

In relative terms, the overall average annual percentage change was
slightly positive -- .71 percent. The weighted averages showed a small
gain by females in hourly earnings relative to males. The average annual
percentage change in relative hourly earnings was negative in only two of
the 11 occupations. The most noteworthy average percentage decrease was
in Clerical and Kindred workers with a .89 percent decrease. The other
average percentage decrease was in Farm Labor with a sezeable 3.03 percent
decrease. In contrast, the largest positive average percentage change in
relative hourly earnings was in Farmers and Farm Managers with a 9.81 per-
cent increase. Other marked average percentage increases were in Nonfarm
Labor with a .36 percent increase; Service workers, except Private House-
hold, with 2.69 increase; and Managers, Officials, and Proprietors with a
1.21 percent increase. Only possibilities (I) and (3) are left: the
relative quantity of labor supplied by males to that supplied by females
is decreasing, while the relative hourly earnings between females and males

is increasing, or remaining relatively constant.

51

2.3.c Wage Rates binndustry

Looking back at Table 7 shows that the overall average percentage change
in the hourly wage rate was 5.27 for males and 5.32 for females. The aver-
age increases are comparable to those found using occupational data. The
industrial estimates are slightly above the occupational ones despite the
fact that relative hourly earnings are less than the relative hourly wage
rates found in Table 10. The largest average percentage increase in the
hourly wage rate for males was in Public Administration with a 6.98 percent
annual increase. Sizeable average increases for males also were in Trans-
portation, Communication, and Other Public Utilities with 5.94 percent and
in Wholesale and Retail Trade with 5.77 percent. Although females exper-
ienced average percentage increases in each of these industries, the average
increases were smaller with 5.66 percent, 5.53 percent, and 4.48 percent
respectively. Females experienced greater average percentage increases
than males in three of the seven industries. The increases were in Agri-
culture, Forestries and Fisheries, and Mining with the greatest difference
of 2 percent; Manufacturing with a .25 percent difference; and Service
workers, except Private Household workers with a .70 percent difference.
2.3.d Relative Wage Rates by Industry

The significance of the figures presented in Table 7 are seen by com—
paring them with those for the relative wage rates in Table 8. The over-
all average percentage change for all industries was negative -— a .17 per-
cent annual decrease. This overall average decrease was a result of average
decreases in four of the seven industries which contained at least 30 per-
cent of the female labor force in 1971. Table 8 shows that Construction
had the largest average decrease in the relative wage rate with a 4.53 per-

cent annual decrease; Public Administration also had a sizeable decrease of

52

1.32 percent; Wholesale and Retail Trade registered a .83 percent average
decrease; and Transportation, Communication, and Other Public Utilities
had a .57 percent average decrease. Most of these decreases were offset
in the aggregate by the .77 percent average increase that occurred in
Service and Finance which employs 48 percent of the female labor force.
The average increase in this occupation, however, was not enough to offset
all of the relative losses.

In summary, the industrial data suggest possibilities (l) or (4): the
relative quantity of labor supplied by males to that supplied by females is
decreasing, while the relative hourly wage between females and males is
decreasing or remaining relatively constant. These suggestions are not
incompatible with those from the occupational data. The common possibility
of (1) between the occupational data and the industrial data leads to the
suggestion that the relative gains made by females with respect to self—
employment income must positively offset the losses incurred by females
working for someone else. This stands to reason, since one is less likely
to cheat, or discriminate against, one's self.

At this point it would be helpful to see how the Women's Bureau inter-
prets the earnings gap and whether or not their study confirms or contra-
dicts these findings.

2.4 Women's Bureau Data and Their Updates

 

The information contained in the Handbook of Women Workers, 1969, per-
tains only to occupations but was obtained from the CPS as was the income
data in this study. It was easy then to take the Women's Bureau's infor-
mation and update it for comparison as well as to construct comparable
figures for industries. The original Women's Bureau study covers the years

1956 to 1966. In reference to this period, they note a widening gap between

53

the median annual wage or salary income of females and males; to be specific,
"a comparison of median wage or salary income of full-time year-round

women and men workers reveals not only that the income of women are consi-
derably less than those of men but also that the gap has widened in recent
years [20, U.S. Dept. of Labor, p. 133]. In 1956, the median wage or

salary income of full-time year-round women was 63 percent of men's and

only 58 percent in 1966. Continuing the estimates to 1971 gives similar
results. The gap between aggregate female and male earnings seems to have
stabilized although it is still greater than it was in 1956. The estimates
were obtained by taking a weighted average of the median wage or salary

incame. The overlapping years are identical. Table 11 presents the results.

TABLE 11

The Relative Aggregate Median Annual Wage or Salary
Income of Full-Time Year-Round Females and Males,
WSFTE/WSFTIE, From 1962 to 1971

 

t t
Year WSFTf/WSFTm
1962 .58
1963 .58
1964 .58
1965 .58
1966 .58
1967 .59
1968 .58
1969 .58
1970 .60
1971 .60

Table 11 also indicates a sizeable gain in 1970 and 1971. Extended esti-
mates of the relative annual wage or salary income between females and males
for all workers and not just the small minority who work full-time year-
round in Table 12 shows the same behavior, but with a wider gap. The sta-

bility in the gap between female annual wage or salary and that of males

54

TABLE 12

The Relative Aggregate Annual Median Wage or Salary Income
of All Werkers Between Females and Males (WSE/ng)
Across A11 Occupations from 1962 to 1971

t t
Year WSf/wsm
1962 .47
1963 .46
1964 .47
1965 .47
1966 .47
1967 .48
1968 .48
1969 .48
1970 .49
1971 .50

can be explained using the information of the previous section; that is,
females are now working shorter weeks more weeks per year for a relatively
lower wage. This is not necessarily inconsistent with the occupational data,
since that data contained self-employment income. This strengthens the
overall case for possibility (3).

The women's Bureau made "A Comparison of wage or salary income of full-
time year-round women workers in selected occupation groups with that of
men [which showed] that women's relative income positions deteriorated in
most occupations during the period of 1956 to 1966" [20, U.S. Dept. of
Labor, p. 134].in the following five occupations:

(1) Professional, Technical, and Kindred workers,

(2) Managers, Officials, and Proprietors, except farm workers,

(3) Sales workers,

(4) Operatives and Kindred workers, and

(5) Service workers, except Private Household workers.
In only the first and the last occupations did the relative income position

of females to that of males improve from1956 to 1966. The update of these

55

categories in Table M of Appendix A, however, indicates that the relative
income position of females improved very little if at all in Professional,
Technical, and Kindred workers; definitely deteriorated from 1962 to 1971
in Managers, Officials, and Proprietors, except Farm; deteriorated also in
Sales workers; improved slightly in Operatives and Kindred workers; and
improved substantially in Service workers, except Private Household workers.
All of the results correspond to the changes that took place as reflected
in Table 6 for all workers, except in Managers, Officials, and Proprietors,
except Farm workers. The difference here is that the occupational date in
Table 6 contained in addition to the wage or salary income of workers their
self-employment income from farm and nonfarm sources. The gains made with
respect to the latter by females positively offsets the losses in the rela-
tive wage between females and males. Although there was a slight gain in
the relative hourly earnings position of females in Sales workers, the gain
was accompanied by a sizeable average percentage decrease in the relative
number of hours supplied by women to that supplied by males. In addition
to the WOmen's Bureau's five occupations the relative hourly earnings posi-
tion of Clerical and Kindred workers also continued to worsen.

An updating for industries reveals a definite upward trend for full-
time year-round workers, but not for all workers as shown in Table 13.

3.0 The Importance of Relative Quantities of Labor Supplied and Relative
Remuneration Rates Between Females and Males

 

 

The evidence presented to now suggests that the most important variable
in explaining the widening gap between aggregate annual female earnings and
male earnings or wage or salary income is the relative quantity of labor
supplied between them. Other possible causes of variation in this gap,

however, were posed in the second question of Chapter 2. One can phrase

56

TABLE 13

The Relative Aggregate Median Annual Wage or Salary Income of
Full-Time Year-Round Workers (WSFTE/WSFTfi) and All Workers
(WSE/Wsfi) For All Females and Males, Industrial Data, 1962 to 1971

 

t t t t
Year WSFTf/WSFTm wsf/wsm
1962 .57 .46
1963 .57 .45
1964 .57 .46
1965 .57 .47
1966 .57 .46
1967 .58 .47
1968 .58 .47
1969 .58 .47
1970 .59 .47
1971 .60 .49

the question as follows: What proportion of the percentage change in the
ratio of aggregate annual earnings or wage or salary income of females to
that of males can be attributed to the percentage change in the relative
number of hours supplied annually by all males and females (Hi/HE) or to
the percentage change in the relative hourly earnings or wage rates between
females and males, (WE/Wi).

Using Equation (2.8) of Chapter 2, the results in Table 14 were obtained.
The calculations vary substantially between occupations and industries. In
the industrial data the variance in the percentage change in WE/W: is 1.81
times the variance in the percentage change in EFt/EMt. The covariance of

t t . . .
the percentage change between Wf/Wh and Hi/H: is 1.71 times the‘varlance

t

t t .
to H /H makes the covari-
f f m

. t t . . . .
in EF /EM , and pOSitive. The inverSion of Hi/H
ance term negative. This makes the relative wage rate and hours comparisdn
. . . . t t . . .
eaSier. POSitive percentage changes in Hf/Hm are assoc1ated Wlth negative
. t t .
percentage changes in Wf/Wm. In general, as the relative number of hours

supplied by all females to that supplied by males increased, the relative

wage rate between the two decreased.

57

TABLE 14
The Relative Importance of the Variance in the Percentage

Change of Hfi/HE, WE/Wfi, and the Cov (H§/HE, WE/Wg)
in Explaining the Variance of EFt/EMt, From 1962 to 1971

Solving Equation (2.8) using industrial disaggregates:

 

Var [A Ln (art/mt) ] = . 00030462 100. 00 percent
Var[ALn(H;/H:)] = .00027141 89.10 Percent
Var [A Ln (wz/win = . 00055221 181 . 28 percent

.00025983 - 170.59 Percent

t t t t
Cov[ALn (Hm/Hf) . A Ln (wt/Wm”

 

99.79 Percent

Solving Equation (2.8) usipg occupational disaggregates:

Var [ALn (art/mat” = .00058494 100.00 Percent

varIALn(H;/H:)] = .00008458 14.46 Percent
t t

VarLALn(Wf/Wfi)] - .00073223 125.18 Percent

CovIALn(H;/H:), ALn(W:/W;)] .00011595 - 39.65 Percent

99.99 Percent

 

In the occupational data, the variance in the percentage change in
WE/W; is 1.25 times the variance in EFt/EMt. The covariance term here is
only a fourth as large as the variance of relative hourly earnings and
positive. The variance in the percentage change in the relative number of
hours supplied between males and females is only .14 percent as large as
the variance in the percentage change in EFt/EMt. The positive covariance
term is interpreted as before. In general as the relative number of hours
supplied by females to that supplied by males increased; that is, Hg/H:
decreased, then the relative hourly earnings between females and males
decreased. Thus, when viewed from both the occupational disaggregation,
and the industrial disaggregation, relative earnings have been less volatile

than relative wages because of the offsetting behavior of relative hours

worked.

58

4.0 Occupational Shifts in the Labor Force

 

Occupational shifts imply a movement of workers among occupations.
Workers may shift from relatively high paying occupations to relatively
low paying occupations or vice versa. The indices discussed in Chapter 2
to capture this movement are presented in two separate sections. First,
the Paasche and Laspeyres indices are presented to show the direction of
the occupational shifts. Then the weighted averages are presented to show
the impact of these shifts.

4.1 Shifts in Occupations

 

The values found for the Laspeyres and Paasche indices of Equations
(2.9) through (2.12) are presented in Table 15. Using the indices three
separate comparisons can be made: (1) male versus female; (2) Laspeyres
versus Paasche; and finally (3) Occupational versus Industrial indices.

In most cases, the industrial female indices are noticeably larger
than the male indices. This means that female hourly wages and annual
earnings have increased at a faster rate than those of males. This does
not necessarily mean, however, that female remunerations have increased
relatively. A previous section of thischapter showed that with respect
to wage or salary income females in general are incurring a slight relative
loss. For occupational data, the female indices are not noticeably larger
than the male indices and in some cases, especially after 1969, they are
smaller. This seems to be the case with respect to both hourly earnings
and annual earnings.

The comparison of the Laspeyres indices with Paasche indices shows
that for males there was little if any occupational shifting, while for
females substantial shifting occurred for wage or salary income. In 1965,

females moved to relatively lower paying industries, since the Paasche

59

TABLE 15
Laspeyres and Paasche Indices of Hourly

Remunerations, Annual Remunerations by Sex
for Occupations and Industries From 1962 to 1971

Industrial Indices

 

Laspeyres Paasche

 

Annual Wage or Annual Wage or

 

 

 

 

 

 

 

 

 

 

 

Wage Rate Salary Income Wage Rate Salary Income

Year Male Female Male Female Male Female Male Female
1962 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
1963 1.05 1.01 1.04 1.02 1.05 1.01 1.04 1.02
1964 1.07 1.09 1.08 1.08 1.07 1.09 1.08 1.08
1965 1.11 1.46 1.14 1.15 1.11 1.38 1.14 1.15
1966 1.16 1.49 1.21 1.20 1.16 1.48 1.21 1.20
1967 1.21 1.49 1.27 1.28 1.21 1.54 1.27 1.27
1968 1.31 1.58 1.35 1.38 1.31 1.63 1.35 1.37
1969 1.44 1.70 1.47 1.46 1.44 1.77 1.46 1.46
1970 1.58 1.93 1.55 1.55 1.58 2.04 1.55 1.54
1971 1.62 1.94 1.60 1.68 1.62 2.06 1.60 1.68

Occupational Indices
Laspeyres Paasche

Annual Wage or Annual Wage or

Wage Rate Salary Income Wage Rate Salary Income

Year Male Female Male Female Male Female Male Female
1962 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
1963 1.04 1.04 1.07 1.03 1.04 1.04 1.05 1.04
1964 1.07 1.10 1.08 1.08 1.07 1.10 1.08 1.08
1965 1.09 1.16 1.12 1.15 1.09 1.15 1.17 1.15
1966 1.16 1.16 1.20 1.15 1.16 1.16 1.19 1.15
1967 1.22 1.24 1.25 1.22 1.22 1.23 1.25 1.22
1968 1.30 1.34 1.32 1.30 1.30 1.33 1.32 1.29
1969 1.42 1.38 1.42 1.34 1.42 1.39 1.42 1.34
1970 1.51 1.50 1.49 1.43 1.51 1.51 1.49 1.44
1971 1.60 1.58 1.56 1.51 1.59 1.59 1.55 1.52

60'

index for hourly wages is lower than the Laspeyres. After 1966, the female
Paasche hourly wage index is substantially above the Laspeyres,indicating
that females moved into relatively higher paying industries. Both male
and female annual wage or salary indices show very little deviation from
each other. In the occupational data, there were very little change and
where changes did occur the female Paasche indices were only slightly
greater than the Laspeyres indices.

The occupational indices were in general less than the industrial
indices. The male indices were very close. On the other hand, there was
a substantial difference between female occupational indices and industrial
indices. This may indicate the females as a whole are making greater gains
with respect to wage or salary income than they are in self-employment
income.
4.2 The Impact of Occupational Shifts

The occupational shifting of females did affect the relative hourly
and annual remunerations between females and males as shown in Table 16.
The estimates were derived using Equation (2.13). The results are presented
separately for occupations and industries. The first three columns of each
section contain annual earnings or wage or salary income. The last three
columns are hourly remunerations. The first column in each set contains
the current values. The second column has the estimates of the respective
relatives with employment held constant. The third column was obtained
by holding annual or hourly remunerations constant and letting the quantity
of labor vary.

The occupational data for both annual earnings and hourly earnings
show that the relative annual earnings and relative hourly earnings between

females and males would have been consistently less if employment patterns

Year

1962
1963
1964
1965
1966
1967
1968
1969
1970
1971

Year

1962
1963
1964
1965
1966
1967
1968
1969
1970
1971

61

TABLE 16

A Weighted Average of Relative Annual and Hourly
Remunerations Between Females and Males for All
Occupations and Industries, From 1962 to 1971

Annual Remunerations

.43
.42
.43
.45
.43
.44
.44
.42
.44
.45

Employment Earnings
Current Constant Constant

.43
.42
.43
.44
.41
.42
.42
.40
.41
.41

.43
.43
.43
.44
.44
.45
.45
.45
.45
.45

Wage or

Employment Salaries
Current Constant Constant

 

.46
.45
.46
.47
.46
.47
.47
.47
.47
.49

.46
.46
.47
.47
.46
.47
.47
.46
.47
.49

.46
.46
.46
.46
.47
.47
.47
.47
.47
.46

Hourly Remunerations

 

Hours

Hourly
Earnings

Current Constant Constant

 

.66
.67
.68
.70
.67
.68
.69
.66
.68
.68

.66
.66
.67
.70
.65
.66
.67
.64
.65
.65

Hours

.66
.66
.66
.67
.68
.68
.68
.68
.68
.69

Wage Rate

Current Constant Constant

 

.75
.73
.76
.77
.76
.76
.77
.75
.75
.76

.75
.72
.76
.77
.76
.76
.77
.75
.76
.76

.75
.75
.75
.75
.74
.75
.75
.75
.74
.74

62

and the number of hours worked had remained as they were in 1962. Allow—
ing employment and the number of hours worked to vary, however, while holding
annual earnings and hourly earnings at their 1962 levels would have made the
relative annual earnings generally larger and the relative hourly earnings
about the same. This implies that the occupational shifting into higher
paying occupations by females has increased their relative annual earnings.
This increase, however, is not as great as it could have been if the in-
creased supply of workers had not lowered the annual earnings in these
occupations.

The impact of the structural shifts in the industrial data was slight.
The relative annual wage or salary income was not affected in general by
shifts in employment from lower paying to higher paying occupations. This
tends to augment the findings in Section 3.0 of this chapter that the co-
variance term between hours worked and wages was an important factor in
explaining the variance in the aggregate annual wage or salary gap between
females and males. The positive covariance term indicated that substantial
increases in the supply of females would be associated with a substantial

decrease in their relative wage.

5.0 Summary and Conclusions

 

The relative number of hours supplied by males to that of females
was found to be decreasing; partly as a result of changing work patterns,
but mostly as a result of the increased labor force participation of fe-
males. The overall direction of change in the relative hourly earnings
between females and males is positive, but the evidence is not strong.
The reason for the doubt is the fact that the relative wage rate between
females and males was found to be slightly negative. The wage or salary

data used for the industrial disaggregates is a large percentage of annual

63

and hourly earnings used for occupational disaggregates. In this latter
group, it was also feund that the relative hourly earnings position of
approximately 33 percent of the female labor force in 1971 deteriorated,
Clerical and Kindred workers from 1962 to 1971 incurred an annual average
percentage decrease of .89 percent in their relative hourly earnings.
These losses, however, were offset by the average annual percentage in-
crease in the relative hourly earnings of 2.69 percent for females in
Service, except Private Household workers. In 1971 approximately 18 per—
cent of the female labor force was situated in this occupation. The
average annual percentage increase in the number of females employed in
the former occupation was 3.39 percent and that of the latter was only
1.16 percent. This would tend to suggest that at some point in the future,
if relative gains and losses continue as over the 1962-71 period, the
average annual percentage change in the relative hourly earnings between
females and males may become negative. For now, however, it is accepted

as positive.

CHAPTER 4

AGGREGATE TRENDS AND THE IMPORTANCE OF RELATIVE HOURLY EARNINGS AND
HOURS WORKED BETWEEN HUSBANDS AND WIVES IN THE ECONOMY

1.0 Introduction

 

This chapter provides an empirical foundation to examine three related
questions. First, what are the recent trends in: (a) the ratio of aggre-
gate wives' earnings to that of their husbands; (b) the relative hourly
earnings between wives and husbands; and (c) the relative quantities of
labor supplied between wives and husbands? Secondly, what proportion of
the gap between the aggregate earnings of wives to that of their husbands'
is attributable to the other trends -- changing relative hourly earnings
between spouses and/or relative quantities of labor which they supplied.
Finally, in the event that changing relative hourly earnings and/or rela—
tive quantities of labor supplied are contributing substantially to the
earnings gap between husbands and wives, what proportion of these changes

are reflections of occupation shifts in their work?

2.0 Aggregate Earnings of A11 Husband and Wife Families

Aggregate earnings of all husband and wife families (EFAM) are iden-
tically equal to the aggregate annual earnings of all husbands (EH) and
the aggregate annual earnings of all wives (EW). Furthermore, the aggre-
gate annual earnings of either all husbands or wives is equal to the pro—
duct of hourly earnings (W) and the average number of hours worked per

year (H) for all husbands and wives; algebraically,

64

65

t _ t t t t
EFAM - Wh Hh + Ww Hw , (4.1)

where husbands are represented by the subscript h; wives by the subscript w;

and the time period by the superscript t. Dividing both sides of Equation

(4.1) by the aggregate earnings of all husbands;

t t _ t t t t
EFAM /EH - l + [(Ww Hw)/(Wh Hh)]' (4.2)
O I I t t
rearranging and solv1ng Equation (4.2) for Wfi/Wh:
t t _ t t t t
Ww/Wh — [(EFAM /EH ) llth/Hw]. (4.3)

. . t _ . .
Substituting EFAM : EFt + EHt into Equation (4.3):

tt: tttt
Ww/wh _ [(EW /EH )(Hh/Hw)]. (4.4)

Equation (4.4) contains all three of the desired ratios (EWt/EHt), the
ratio of aggregate wives' earnings to that of their husbands'; (HE/H5),
the relative number of hours worked per year or the relative quantities of
labor supplied by both husbands and wives; and (Wt/WE), the overall rela-
tive hourly earnings between all wives and husbands.

Annual estimates for the median EWt and the median EHt are easily con-
structed from aggregate income data, but following the procedure in Chap-
ter 3, a weighted average is used to correct for data insufficiencies.
Estimates for either Hfi/H: and WS/Wfi, however, are not easily accessible
since they have to be constructed from the disaggregated occupational data
as in Chapter 2. Again disaggregate hourly earnings for husbands and wives
do not exist as was true for all males and females. Information on the
average number of hours worked per year by spouse is derived by using the
average number of hours worked per year by males and females in a particular
occupation. The total number of hours worked per year by all husbands and

all wives is equal to the product of this occupational proxy for the average

66

number of hours worked per year by males or females and thernmber of hus-
bands or wives in that occupation. Aggregating across all occupations

by spouse gives the total number of hours supplied by husbands, HE; and the
total number of hours supplied by wives, H5. The ratio of H: to H: is

expressed as follows:

 

n
2 .HPYt - .Qt
._ i h i h
t t i—l
Hh/H = ,
w n t t
Z .HPY ' .Qw
. i w 1
i=1

where H the total number of hours supplied,

HPY = the average number of hours worked per year approximated
from the occupational data,

Q = the total number of workers employed,
t = the year,

h = husbands,

w = wives,

i = the ith occupation,

n 11.
By substituting Equation (4.5) into Equation (4.4), the aggregate annual
relative hourly earnings between wives and husbands, Ws/Wfi, can be deter-
mined for all husband and wife families in the economy. With the appro-
priate data for 1962 through 1971, all three ratios in the first question
of this chapter can be examined.

Minor modifications of Equation (4.4) are necessary in order to exa-
mine the second question: What proportion of the gap between aggregate
wives' earnings and that of their husbands is attributable to the other
trends -- changing relative hourly earnings or hours worked between hus-

bands and wives. Taking the first difference of the natural log and

rearranging:

67

t t _ t t _ t t
ALn(EW /EH ) — ALn(Ww/Wh) ALn(Hh/Hw). (4.6)

Equation (4.6) now gives the percentage change in each one of the ratios --
EWt/EHt t/Ht and Wt t B findin the varian e ' h t 't i

, Hh w' w/wh° y g c in eac erm, i 5
possible to determine how much of the variance in the percentage change in
the earnings gap between wives and husbands is explained by the variances
of the percentage changes in the other two terms on the right-hand side of
Equation (4.6).

t t t t t t
VarlALn(EW /EH )1 - Var[ALn(Ww/Wh)] + Var [ALn(H.h/Hw)]
t t t t
- 2Cov[ALn(Ww/Wh) , ALn(Hh/Hw) ] , (4.7)

then dividing each side of Equation (4.7) by the left—hand term;

t t t t
1 = Var [ALn(Ww/Wh)] Var [ALn(Hh/Hw)]

 

 

Var [A Ln (EWt/EHt) I Var [A In (mt/Eat) ]

-2COV[ALn (wt/WE) , ALn (HE/H5”

 

. (4.8)
Var [ALn (mt/Eat”

Equation (4.8) indicates how much of the variance in the percentage change
in the ratio of aggregate wives' earnings to that of their husbands is
explained by the variance in the percentage change in the relative hourly
earnings between wives and husbands, by the variance of the percentage
change in the relative quantities of labor supplied by both spouses, and
the covariance between their relative hourly earnings and quantities of
labor supplied. The development of Equation (4.8) should indicate which
forces -- relative hourly earnings, relative quantities of labor supplied,
or their interaction -- are having the greatest impact upon the percentage
change in EWt/EHt.
Finally, given that relative hourly earnings between wives and husbands

and/or their respective quantities of labor supplied are influencing the

earnings gap between wives and husbands; indices need to be constructed

68

as in Chapter 2 for all husbands and wives to measure the direction and
impact of occupational shifts. Unlike the previous chapter, there are
only two Laspeyres indices and two Paasche indices. The two Laspeyres
indices are a husbands' hourly earnings index and a wives' hourly earnings
index; similarly for the Paasche indices. If the Paasche indices are very
close to the Laspeyres indices then very little occupational shifting is
occurring. On the other hand, if the Paasche indices are greater than the
Laspeyres indices, then husbands and/or wives are moving into higher paying
occupations.

A weighted average of the wives' hourly earnings and that of their
husbands shows the impact of any occupational shifting. Taking the ratio

of the two weighted averages as follows:

n

 

 

 

2: .wt - .Ht
. i w i w
i=1
n
t z .Ht
w'w i=1l w
—'t'= n r ' (409)
W t t
h .2 iwh uh
i=1
n
t
.2 iHh
i=1

('1'

where Wi/W the relative wage between all wives and husbands,

S 5
II

the wage rate,

£
ll

wives,
h = husbands,
t = the year,
i = the ith occupation,
n = 11.
Equation (4.9) is calculated using current hours and hourly earnings,

then it is recalculated first using current hourly earnings and base year

69

hours and then using base year hourly earnings and current hours. The
difference in estimates of relative hourly earnings between wives and
husbands using current hourly earnings and hours worked and some combina-
tion of base year hourly earnings or hours worked shows the impact of occu-

pational shifting by husbands and wives upon WE/Wfi.

3.0 The Data - Sources and Alterations

 

3.1 Sources

The data necessary for the models presented came from two sources.
First, hourly earnings and hours worked are taken from Chapter 3. As pre-
viously mentioned the number of hours worked per year by a wife is approxi-
mated by the average number of hours worked per year per female in her
occupation. Correspondingly, the husband's hourly earnings and hours
worked are approximated. Second, the marital labor force participation
rates, EENt, are derived from the number of husbands and wives in a par-
ticular marital labor force participation pairing. This information was
found in "Marital and Family Characteristics of WOrkers," another annual
Special Labor Force Report published by the Department of Labor/Bureau of
Labor Statistics [18]. The data, like those in Chapter 2, were originally
obtained from the Current Population Survey. Two tables -- "Occupation
Groups of Employed Married Women, Husband Present by Employment Status
and Occupation Group of Husband, March 19 ," and "Employment Status of
Married Women, Husband Present, By Employment Status and Occupation Group
of Husband, March 19 " -- were combined to encompass the employment
statuses of all husband and wife families in the U.S. from 1962 to 1971.

3.2 Alterations

 

Given the data above a husband and/or his wife may be employed, unem-

ployed, or not in the labor force. If they are employed, each falls within

70
one of the 11 occupations. If either is unemployed or not in the labor
force, he or she is lumped into one category: not in the paid labor force.
Those husbands and/or wives in this category have zero earnings. Thus
with 11 occupational categories and one for not in the paid labor force,
there are 144 possible marital labor force participation pairings, each
of which has a marital labor force participation rate.

4.0 Relative Hourly Earnings and Hours Wbrked Between All Husbands and
Wives in the Economy

 

The ratio of aggregate wives' earnings to that of their husbands in-

creased over the period of 1962 to 1971. The annual estimates for EWt/EHt

rose from .17 in 1962 to .21 in 1971 as shown in Table 17. The average

TABLE 17

The Ratio of Aggregate Wives' Earnings to
That of Their Husbands', EWt/EHt, For All
Husband and Wife Families, From 1962 to 1971

Year EWt/EHt
1962 .17
1963 .17
1964 .17
1965 .18
1966 .17
1967‘ .18
1968 .19
1969 .19
1970 .20
1971 .21

. . t t .

annual percentage increase in EW /EH was 2.47 percent. Comparing these

. t t t t . . .
estimates of EW /EH to EF /EM in Table 3 showed that working wives earned
a smaller proportion of aggregate family earnings, than all females earn
of aggregate earnings. Although the initial difference between the esti—

t t t t . .

mates of EW /EH and EF /EM in 1962 was approx1mately 3 percent, by 1971

the difference had widened to approximately 6 percent.

71

Considering just those families in which both the husband and wife
worked a substantially different estimate of EWt/EHt occurred. Table 18
contains the estimates for this subpopulation of all husband and wife

families. For this particular subpopulation, the annual estimates of

TABLE 18

The Ratio of Aggregate Wives' Earnings to That of
Their Husbands', EWt/EH , for Families in Which
Both Husband and Wife work, From 1962 to 1971

Year EWt/EHt
1962 .45
1963 .44
1964 .45
1965 .47
1966 .43
1967 .44
1968 .45
1969 .43
1970 .44
1971 .44

EWt/EHt remained relatively constant from 1962 with .45 percent to .44
percent in 1971. Comparing these estimates of EWt/EHt to EFt/EMt in
Table 3, it can be seen that working wives earn a larger proportion of
aggregate family earnings, than all females earn of aggregate earnings.
In contrast to the estimate of EWt/EHt for all husband and wife families,
the gap between EWt/EHt for just those families in which both husbands
and wives worked and EFt/EHt decreased over the period.

Aggregating the number of hours worked by all husbands and comparing
it to the number of hours worked by all wives across the possible marital
labor force participation pairings, showed that the relative quantity of

hours supplied was decreasing. Table 19 shows that all husbands worked

about four to five times the number of hours than did their wives. The

72

continual average annual percentage decrease of 2.06 percent indicates
a disproportionate increase in the relative number of hours supplied by
wives. Again, the literature on labor force participation rates of married

women would anticipate these findings.

TABLE 19
The Relative Number of Hours Supplied by Husbands To That
Supplied by Wives, Hfi/Hs, for All Husband and Wife Families
and For Those in Which Both Werked, From 1962 to 1971

All Husband and Families in Which

 

Year Wife Families Both Worked
1962 5.38 1.97
1963 5.31 2.01
1964 5.24 2.02
1965 5.13 1.99
1966 5.14 2.05
1967 4.93 2.03
1968 4.77 2.06
1969 4.66 2.07
1970 4.57 2.10
1971 4.47 2.08

In contrast, Table 19 also shows that the relative number of hours
supplied in the subpopulation of families in which both worked increased
slightly. In addition this ratio, Hfi/HE, was slightly below that for the
entire labor force, Hi/Hz, found in Table 4. This is probably because
single men work fewer hours than married men.

Using the appropriate estimates for EWt/EHt in Tables 17 and 18 and
those for HE/H: in Table 19, Equation (4.4) is solved for wi/Wfi. Table 20
presents the results for all husband and wife families and for those
families in which both worked. The relative hourly earnings for all wives
and husbands were consistently higher for families in which both worked

than in all husband and wife families. The difference, however, was slight.

The average annual percentage change in the latter group was only .41 percent.

73

TABLE 20
Relative Hourly Earnings Between Wives and Husbands,
Wfi/Wfi, for All Husband and Wife Families and for Those
in Which Both werked, From 1962 to 1971

A11 Husband and Families in Which

 

 

 

Year Wife Families Both worked
1962 .89 .89
1963 .89 .89
1964 .90 .90
1965 .93 .93
1966 .89 .89
1967 .90 .90
1968 .92 .92
1969 .89 .89
1970 .91 .92
1971 .92 .93

The percentage changes in WE/W: for both groups followed the same pattern.
This demonstrates that the association between the wife's hourly earnings
and that of her husband's is not only a positive one as Mincer notes, but
also a very consistent association.

The question to be posed now is how the percentage change in the ratio
of aggregate wives' earnings to that of their husbands' (EWt/EHt) is asso-
ciated with the percentage changes in the relative number of hours supplied
annually by all husbands to that of their wives (Hi/H5), or to the percen-
tage changes in the relative hourly earnings between them (WS/Wfi). Using

Equation (4.8), the following results in Table 21 were obtained. The factor

TABLE 21

The Relative tImportance of the Variance in the Percentage
Change of (Hh/Hw),( (Wt/Wh), and the Cov(Hh/Hw, Ww/Wh) in
Explaining the Variance of EW t/EHt, From 1962 to 1971

varlALn(EWt/EHt)] .00117061, 100.00 Percent

VarlALn(Hh/H:)] = .00015733, 13.44 Percent

VaerLn(W:/W E)] = .00070130, 59.91 Percent
t t t t

Cov[ALn(Hh/Hw), Ln(Ww/Wh)] = .00015975, 27.29 Percent

74

contributing the most variance in the percentage change in the ratio of
aggregate wives' earnings to that of their husband was the variance in

the percentage change in the relative hourly earnings between all wives

and husbands. The variance in the percentage change in WE/W: is 60 percent
as large as the variance in the percentage change of EWt/EHt. The second
largest factor was the covariance in the percentage change between Hfi/H:
and Wt/W: which was 27 percent as large as the variance in the percentage
change of EWt/EHt. As in the model for all males and females, the variance
in the percentage change in Hfi/H: amounts to only 13 percent of the variance
in the percentage change in EWt/EHt. All of these findings are very simi-
lar to those for the entire labor force.

Unlike the covariance term found for all males and females in the
labor force, as previously presented in Table 14, the covariance term for
all husband and wife families was negative. This is interpreted to mean
that if the percentage change in Hfi/H: is positive, then the percentage
change in Wt/Wt is negative. Inverting HIE/Ht to Ht/Ht makes the covariance

w h w w h
term positive, which, in turn, means that if the percentage change in

t . . .
is also pOSitive.

t t . . . . t
Hw/Hh is pOSitive, then the percentage change in ww/Wh

Thus if the relative number of hours supplied by wives to that supplied

by their husbands increases, the relative wage between wives and husbands
increases -- and vice versa. The difference in behavior for all husband
and wife families and the entire labor force with respect to the covariance
term could be the result of changes in supply and demand forces. For all
wives, if the demand for the labor increased, a rise in their relative
hourly earnings and hours worked is expected, and conversely. For all
females, however, as the supply of their labor increases, a decrease in

their relative hourly earnings is expected, and conversely. In short,

75

all wives may be moving along their supply curve while all females may
be shifting their supply curve.

Whether or not there has been any appreciable movement over the period
among occupations by husbands and/or their wives can be surmized from the
Laspeyres and Paasche indices. The Paasche indices in Table 22 are slightly
smaller than the Laspeyres indices indicating that husbands and wives are

moving into relatively lower paying occupations. This may indeed be the

TABLE 22

Laspeyres and Paasche Hourly Earnings Indices
for All Wives and A11 Husbands, From 1962 to 1971

Laspeyres Paasche
Year Husbands Wives Husbands Wives
1962 1.00 1.00 1.00 1.00
1963 1.04 1.06 1.04 1.06
1964 1.07 1.11 1.07 1.11
1965 1.09 1.17 1.09 1.16
1966 1.16 1.19 1.15 1.18
1967 1.22 1.26 1.21 1.25
1968 1.30 1.38 1.29 1.38
1969 1.43 1.47 1.42 1.46
1970 1.53 1.61 1.52 1.60
1971 1.62 1.71 1.61 1.71

case, since one out of every three married women work in Clerical and
Kindred workers, which was shown in Chapter 3 to be an occupation where
females in general were incurring relative hourly earnings losses.
Table 23 contains the weighted averages to show the impact of the
shifting. By wives staying in their 1962 employment patterns, and the
hourly earnings varying, the relative hourly earnings between all wives
and husbands would have been greater than found by using current hourly
earnings and hours worked. On the other hand, if hourly earnings were

held constant while employment varied, then the relative hourly earnings

76

between all wives and husbands would have been smaller than those in

current terms.

TABLE 23

Weighted Averages of the Relative Hourly Earnings
Between All Wives and Husbands, From 1962 to 1971

 

 

 

Current Hourly Current Hourly Base Hourly Earn—
Earnings and Earnings and Base ings and Current

Year Hours worked Hours Worked Hours Werked

1962 .89 .89 .89

1963 » .89 .91 .87

1964 .90 .93 .87

1965 .93 .95 .87

1966 .89 .91 .86

1967 .90 .92 .87

1968 .92 .95 .86

1969 .89 .91 .87

1970 .92 .94 .87

1971 .92 .94 .87

This tends to indicate that if wives had not incurred the relative
losses of all females in hourly earnings in Clerical and Kindred workers,
the relative hourly earnings between them and their husbands would have
Ibeen greater. At the same time, the larger proportion of working wives
in Professional, Technical, and Kindred workers; in Managers, Officials,
.and Proprietors; and in Operatives and Kindred workers has kept relative
liourly earnings between wives and husbands from falling even more. The
1:hird column of Table 23 indicates how low relative hourly earnings between
vwives and husbands would have gone if females as a whole had not made rela-
tive hourly earnings gains in the above mentioned occupations. As a matter
of reference, Table 24 contains the occupational distribution of the female

and Inale labor forces and that of all wives and husbands.

5-0 .A Conditional Probability Distribution of Husbands and Wives
'The stability of the consistent association between what a wife can

earn_ On an hourly basis as compared to that of her husband depends on the

77

TABLE 24

The Occupational Distribution of All Females, All
WOrking Wives, All Males and All working Husbands In
1971 By Percentage of Totals in Each

 

Occupation All Females A11 Wives All Males All Husbands
Professional 13.78 15.87 12.95 14.37
Managers 4.02 5.23 12.01 17.65
Clerical 32.95 33.16 6.70 6.06
Sales 7.47 7.43 6.00 6.06
Craftsmen 1.18 1.20 20.21 21.55
Operatives 13.42 14.48 18.59 17.75
Private Household 5.16 3.41 .07 2.34
Service 18.07 16.80 8.78 4.42
Laborers .87 .91 9.00 5.21
Farmers .47 .17 3.10 3.56
Farm Laborers 2.61 1.27 2.58 .96

 

 

78

socioeconomic factors, somewhat divorced from the market. If Pij represents
the conditional probability that a wife is in the jth occupation, given

that her husband is in the ith occupation, then a two-way analysis of
variance model can be used to test which factors determine Pij’ The model
is as fellows:

= 11+ 01 + Bj + Yi‘ + 8.. (4.10)

Pijk 3 13k'

Pijk is first a positive function of u, the overall mean for all husband
and wife families. This mean is the probability that any given husband
and wife will find themselves in a particular marital labor force partici-
pation pairing, 13. To this probability is added the probability that any
husband is in the ith occupation, ai. Correspondingly, Bj represents the
probability that any wife may find herself in the jth occupation. ai is
considered the fixed row effect, while Bj is the fixed column effect. To
these terms, an interaction term, Yij' is added. Yij represents the pro-
bability that given the husband's occupation, the family will be in any

of the remaining labor force participation pairings, ij. Finally, Eijk

is the error term. The conditions of Equation (4.10) are the following:

i

12

1) 2 di = 0,
i=1
12

2) Z 8. = O,
j=1 3
12 12

and 3) Z Z Yi' = 0.

i=1 j=l 3

The above analysis of variance model involves only two assumptions. First,
the column and row effects are nonadditive, the reason for the interaction
term. Secondly, the Eijk's are independent random variables, normally

. . . . 2
distributed with mean 0 and variance 0 . They represent the extent to

79

which the data depart from the model because of random disturbances. The
significance of the ai's, Bj's, and yij's are tested with the F-statistic.

After combining the two BLS data sets to derive the marital labor
force participation rates, the conditional probabilities were calculated
and are summarized in Table 25. These probabilities show that husbands
in farm related occupations have the highest probabilities of having a
non-working wife. For husbands in Farmers and Farm Managers, moreover,
the probability that their wives, if they worked, were also in farm re-
lated occupations was a combined total of 14.3 percent. At the other end
of the rank ordering, husbands in Professional, Technical, and Kindred
workers had a 15.1 percent chance that their wives, if they worked, were
in the same occupation. Despite the occupations of most husbands, there
was always a good chance that a working wife was in Clerical and Kindred
workers. This suggests that wives tend to be in occupations related to
those of their husbands. For example, if their husbands are employed in
white collar occupations then there is a greater probability that wives,
if employed, are in related white collar occupations than in blue collar
occupations and vice versa.

The analysis of variance model represented by Equation (4.10) con-

firms these suspicions. The results were as follows:

 

 

Source of Sum of Degrees of F- Level of
Variance Sguares Freedom Statistic Significance
i .000357 11 .1952311 .998
j 41.516239 11 22700.87 <.0005
ij 1.411839 121 70.18058 <.0005
.. .215471 1296
ijk

Total 43.143906 1439

80

 

momom Honda

 

b.5h H. H. H. m.v m.N v.v m. m.H h.v m.H m.N onm ecu CH uoz
m.Nh m.m v. m. H.m o.m N.v H. o.H N.m N. m.H mumuonmq Show
v.00 m.mH o.H H. v.m o.H 0.4 v. v.H m.m H.H m.m mnoeumm
m.mo m. N. m. N.m o.m b.m m. h.H H.h m. h.H muoHOQMA
o.mm H. o. N. o.mH m.v m.m m. m.N m.m v.H v.m mofl>umm

. . . . . . . . . . . . oHonmm5om
m mm H o m o m H H m m m m m H 0H m H w v mum>Hum
m.Hm m. o. m. «.5 o.m «.HH c.m m.m m.OH H.H v.m mmeeumumao
h.m© N. H. N. N.@ m.H m.h o. m.N m.NH v.H v.m coEmummnu
m.om o. o. H. o.v m. h.N v. m.v m.oH v.N m.h monm
H.mm H. o. H. m.m o.H m.m w. N.m b.mH o.N m.m HmoHuoHO
m.Nm H. o. H. v.m e. o.N v. @.v m.mH v.v N.© mnommcmz
h.vw H. H. o. m.N m. m.H m. H.N v.NH m.H H.mH Hmconmwwonm
m. e a e m... .. a. a n n n a. m w. n 8.8...
1 1.1. m.m m x n I. a e T. e u 0 mo
m p r. 1 e 1 m. a m u v... m u. m a.

g.u a 1 a o u.3_ 4 o a s coHumm

m. .4 m s m a m. a I. m e 1 s I5ooO

u. p m u I s m. 85.
1 a s w mo
I coHumm5ooo

 

 

HhmH ou NomH .mmHHHamm wMHz can ocmnm5m HHd
mo soHu5nHHumHo HHHHHnmnoum HchHuHocou 0:9

mm mqm¢9

81

The ai's were insignificant. The husband's occupation had no direct
influence on the conditional probability that a wife was in the ith occu-
pation, given that of her husband. The Bj's, however, were significant.
This is interpreted to mean that the distribution of working wives as
found in Table 25 is not by chance. The interaction term, Yij’ was also
significant,meaning that the husband's occupation places his wife on the
low or high end of the rank ordering of female occupations, the Bj's. As
suggested, white collar husbands tend to have white collar wives and vice
versa. Wt/W: is highly correlated because males and females pick mates
with similar socioeconomic backgrounds, and therefore their occupational

opportunities are very similar.

6.0 Summary and Conclusions

 

In this chapter it was found that the relative number of hours supplied
by all husbands to that supplied by all wives, Hfi/HE, was decreasing. This
result is very similar to that found in Chapter 3 for all males and females.
A different result, however, arises when just families in which both hus-
band and wife worked are considered. The relative number of hours supplied
between husbands and wives was gradually increasing. Thus H:/H: in Equa-
tion (1.5) is steadily decreasing. For this to occur despite the increased
labor force participation of wives, the number of wives taking part-time
jobs must be significant.

Relative hourly earnings between all wives and husbands was found to
have increased slightly. In addition the estimates of between all wives
and husbands were found to be very high -- that is, the wife's hourly
earnings was approximately 90 percent of her husband's. The analysis of

variance model showed that this relationship was relatively stable since

82

the occupational opportunities available to a wife were dependent upon
those opportunities available to females in general and the socioeconomic

characteristics of her husband.

CHAPTER 5

THE CONTRIBUTIONS OF MARRIED WOMEN TO AND THE EFFECT OF
RELATED FACTORS ON THE DISTRIBUTION OF AGGREGATE FAMILY EARNINGS

1.0 Introduction

 

At the beginning of this study an identity as represented by
Equation (1.5) was introduced to show the impact of the increased labor
force participation of married women, the relative hourly earnings be-
tween all wives and husbands, the relative number of hours supplied by
wives and husbands, the relative hourly earnings between all males and
females, and finally, the relative number of hours supplied between all
males and females on the distribution of family earnings. The purpose of
this chapter is to integrate the findings of the previous chapters, to
determine aggregate family earnings and its distribution, and to deter-
mine the impact of the above five factors on this distribution.

In brief, the evidence in Chapter 3 led to the following conclusions.
First the relative number of hours supplied by all males to that supplied
by all females is decreasing. Furthermore, it was found that the relative
hourly earnings between all females and males was increasing slightly.

The industrial wage or salary income data, however, suggests that this
may be only a temporary phenomena, since relative wages between females
and males in the labor force had a negative average annual percentage
change. The proportion of earnings that is wage or salary income is the

predominant factor in drawing this conclusion. Finally, the Paasche and

83

84

Laspeyres indices for males showed very little occupational shifting as
defined as males either moving into relatively higher or lower paid occu-
pations. The impact of the absolute increases indicated by the Paasche
and Laspeyres indices can be assumed equally distributed, thus (WE°H:)
scalarly increased across all occupations.

Chapter 4 considered the related factors that pertained to all hus-
band and wife families. It was found that for those families in which both
husband and wife worked, the relative hourly earnings between all wives
and husbands had a positive average annual percentage change. In contrast
to the entire labor fbrce, the relative number of hours supplied by all
husbands to that supplied by wives in those families in which both worked
was found to have positive average annual percentage changes. The in-
creased labor force participation of married women is an established fact
and needs no further documentation. Finally, it was found using the con-
ditional probability distribution that a wife's hourly earnings is highly
correlated with that of her husband's and that this relationship is rela-
tively stable over time.

Aggregate family earnings were calculated using Equations (1.2),
(1.3), and (1.5). Gini coefficients were calculated for each year (see
Appendix B). Then as outlined in Chapter 1, the increased labor force
participation of married women and the other five related factors were
separately held constant while all else varied to generate six adjusted
Gini coefficients. Two other measures of income equality are used to help
interpret the Gini coefficients. First, to illustrate the subtler changes
that are taking place -— in particular, the extent to which high income
females are working as compared to low income females -- the quintile

distribution of hours worked and earnings received are explored. Second,

85

in order to see if these subtler changes are making families relatively
better or worse off, absolute and relative measures of income equality
are also employed. Finally, if the identity in Equation (1.5) proves to
be adequate then the impact of various related factors and the increased
labor force participation of married women 0n the distribution of

family earning can be determined.

2.0 Aggregate Family Earnings and Its Distribution

 

Aggregate family earnings, AEFAMt, as calculated in Equation (1.3),
are presented in Table 26. Over the ten-year period they have continued
to increase and have almost doubled while the number of husband and wife

families has increased only 4 percent. The Gini coefficient of aggregate

TABLE 26

Aggregate Family Earnings, AEFAMt, and Husband
and Wife Families in Thousands, From 1962 to 1971

Number of Husband

 

Year AEFAMt and Wife Families
1962 196,366,954 40,309
1963 208,689,081 40,836
1964 220,424,357 41,142
1965 232,415,584 41,501
1966 ’ 253,377,545 41,954
1967 271,043,547 42,395
1968 297,684,416 43,057
1969 326,014,396 43,559
1970 347,297,306 44,029
1971 359,055,128 44,445

family earnings as calculated for each year are presented in Table 27.

As a Gini coefficient approaches one, the earnings distribution is be-
coming more disperse. The Gini coefficients in Table 27 range from .31125
in 1968 to .34139 in 1971. Thus, over the ten-year period variation is

relatively small and cyclical in character. The stability of these Gini

86

TABLE 27

Gini Coefficients for Aggregate Family
Earnings, AEFAMt, From 1962 to 1971

 

Year Gini Coefficient
1962 .3249
1963 .3258
1964 .3225
1965 .3179
1966 .3120
1967 .3138
1968 .3113
1969 .3180
1970 .3211

1971 .3414

87

coefficients can be deceiving since identical Gini coefficients can be
derived from two very different Lorenz Curves. The actual plotting of the
underlying Lorenz Curves for selected years helps to detect subtler changes.
Figure 2 contains such plots for 1962, 1966, and 1971.

The Lorenz Curves show that the lower income families gained in their
share of aggregate family earnings between 1962 and 1966, but by 1971 had
lost even more than the original gain. The percentage of aggregate family
earnings received by the higher levels of family earnings continued to
increase over the entire time period. Figure 2 also suggests that the
family earnings level for about 10 percent of all families -- those between
70 and 80 percent of all families -- has a special significance. The
Lorenz Curve pivots in this region. The families that fall before the
focal point have incurred relative losses, while those after this point
seem to have made relative gains over the period.

Dividing the Lorenz Curves into quintiles shows where and how these
gains and losses were incurred. Thus, the horizontal axis of the Lorenz
Curve is partitioned into five equal sections or quintiles. Each section
contains 20 percent of all husband and wife families. Similarly the ver-
tical axis is partitioned such that each quintile of husband and wife
families received their respective amounts of aggregate family earnings.
The underlying disaggregate data used to calculate AEFAMt also gave the
number of hours supplied by each quintile as found in Table 28.

From 1962 to 1971 the percentage of aggregate family earnings received
by Quintiles IV and V increased by about 1 percent each. Quintiles I and
III each lost about .5 percent of aggregate family earnings, while Quin-
tile II lost 1 percent. The gain of the top two quintiles was definitely

at the expense of the bottom three quintiles.

Cumulative Percentage of Aggregate Family Earnings

88

100

 

50"-I

     

’ t

 

 

 

T’EFAM
in
lowest<—-> highest
0 ' I ”I I T ”1* I I
50 100

Cumulative Percentage of Families of A11 for a Particular {SEFAMt

FIGURE 2

The Lorenz Curves of Aggregate Family
Earnings in Selected Years

89

 

 

 

 

 

 

 

 

 

 

 

 

mm.Hm mN.wH Ho.mm mh.Nm NH.nH oo.OOH Hmuoe
mm.mN Hm.MN Nm.m MN.mm mH.mN mv.mN Ho.oH mm.mm mNmnHIHnnHH >
hh.HH Nm.NN vo.m om.mN hH.m vm.mN om.N om.>N onnHHIvvooH >H
v0.5 oo.mH mm.H v0.0N m¢.m oo.mH mm.H No.HN vaOHImnon HHH
mm.mH HN.VH om.m Hm.hH mm.0N MN.HH mm.N mm.mH NnomINHmm HH
on.mH mm.H we. hm.N mm.NN 0N.H mm. mm.H HHmmIo H
HhmH HhmH
mm.vm hm.mH oo.OOH Hm.mm mH.vH oo.OOH H6509
ON.0m av.NN mn.m NN.Nm mm.bN no.mN Hm.m Hm.vm HmBOHIHMOH >
Hh.h mN.vN No.N hN.mN Hm.m mH.mN mm.H on.oN mmonIovom >H
em.v m>.wH mm. on.mH m¢.m mo.HN mm. vm.HN mvothvmv HHH
mm.VH mo.¢H mv.N hm.wH em.MH vm.NH no.N mm.vH ovmVImmmH HH
hm.m Nb.v we. mH.m oo.mH mo.H Hm. oo.N NmmHIo H
NomH NmmH
nut.% Q.H_% q.H % H.% 0.3 % H H % H % % as
n.u H ,A m o .A m o m o .A m a m n o m o o n.
5.... a... we... s... new we... a... I. w.
moanm :50 1
m3 fa sum. 4.... as... mpn pn n .6 m T.
903 so? m sTP sqa QR S I
M:M.M u 1 e 1 e 1 u tn A a
M 1.x m.w.r. w.T. w.T. 1.1. W W W
a u x. o. o. o. nun 2. mm 1. 3
s . a a A
o. a
S
Han can NmmH :H .oHHucH5O 2n omxno3 mH502 one mmchnmm HHHEmh mummonmma

mm mHm<B

90

The percentage of aggregate family earnings received by working wives
in each quintile increased from 1962 to 1971. There was almost a doubling
of the percentage of aggregate family earnings received by wives in Quin-
tile II. The increases reflect almost a 25 percent increase in the percen-
tage of earnings received by all wives. In particular, in 1962 wives
earned 14.20 percent of aggregate family earnings. By 1971, they were
earning 17.12 percent of aggregate family earnings. In contrast, the per-
centage of aggregate family earnings received by husbands decreased from
85.81 percent in 1962 to 82.78 percent in 1971. The percentage of earnings
received by husbands in each quintile decreased in the first three quintiles.
The percentage of aggregate family earnings received by husbands in the
top two quintiles increased.

The quintile distribution of the number of hours worked by all husband
and wife families is much more equal than the distribution of their earn-
ings. From 1962 to 1971, the percentage of aggregate hours supplied by
working wives and husbands decreased substantially in Quintile I. This
percentage increased in Quintiles II and III, decreased in Quintile IV,
and increased in Quintile V. The percentage of aggregate hours worked
by wives in each quintile increased except in Quintile I. In total, wives
increased the total number of hours they supplied by 25 percent, from
15.67 percent in 1962 to 18.28 percent in 1971. Husbands, on the other
hand, showed a decrease in the total number of hours supplied by them
from 84.32 percent in 1961 to 81.33 percent in 1971. This was a result
of percentage decreases in the total number of hours supplied by husbands,
particularly in Quintile I, and in Quintile II. This percentage increased

in the other three.

91

These estimates of hours worked are not necessarily inconsistent
with those found in Table 19. Included in the quintile analysis are those
families in which wives worked and the husband did not, those families in
which the husband worked and the wife did not, and those families where
neither worked. Otherwise, aggregate family earnings would be incomplete.

The contribution of working wives is better illustrated, however, by
looking at what percentage of the earnings received or hours worked by a
particular quintile came from wives. In 1962, working wives contributed
15.66 percent of Quintile I's earnings while supplying 8.87 percent of
the hours worked in that quintile. In Quintile II, they again earned a
sizeable portion of the earnings received by that quintile, 13.84 percent,
and supplied a greater percentage of the hours worked, 14.98 percent. The
percentage of earnings received and hours worked by wives in Quintiles III
and IV were about half as much as the first two quintiles. The largest
proportion earned and hours worked by married women was in Quintile V.
Wives earned close to a third of family earnings in this quintile, 30.20
percent, and supplied 27.35 percent of the hours worked. In 1962 wives in
Quintile V supplied 21 times the number of aggregate hours worked by
married women in Quintile I.

In 1971, the percentage of earnings received and hours supplied by
married women increased in every quintile, except for Quintile V. In
this quintile a negligible decrease occurred in the percentage of hours
worked by wives. The sizeable increase in these percentages in the bottom
two quintiles is worthy of note, yet high income females are still sup-
plying relatively more hours. Wives in Quintile V supply 23 times the
number of aggregate hours worked by wives in Quintile I, a slight in-

crease over 1962's estimates.

92

The impact of the increases in hours worked and earnings received
by the wives of the top two quintiles can be illustrated using absolute
and relative income dispersion measures. First, absolute income disper-

sion is measured by taking the mean annual earnings of Quintile V and sub-

tracting those of Quintile I. The difference is absolute income disper-
sion. In 1962 Absolute income dispersion was $7,898, while in 1971 it
was $13,294. Absolute income dispersion increased by $5,294. But this
piece of information is worthless without a measure of whether or not

Quintile I is now relatively worse off than before. This is measured

by the relative income dispersion, found by taking the mean annual earn-
ings of Quintile V and dividing it by that of Quintile I. In 1962, the
relative income dispersion for all husband and wife families was 8.93;
that is, Quintile V had an average family earnings level 8.93 times that
of Quintile I. The relative income dispersion for 1971 was 8.97. Thus
while absolute income dispersion has almost doubled, the relative income
dispersion has only slightly increased. Proportionately more hours are
supplied by high income wives than by low income wives, making the distri-
bution of family earnings more disperse. The occupational distributions
of married women for 1962 and 1971 are shown in Table 29. The percentages
at the bottom of the table do not add up to 100 percent because many
working wives are miscellaneously scattered throughout the various occu-
pations. At most, 6 percent of the wives in any one quintile have not
been accounted. Table 29 indicates that if a husband had a working wife,
the chances of their family earnings level being within the boundaries

of Quintile V were very good. The percentage of working wives has in-
creased in every quintile from 1962 to 1971. In Quintile I most wives

do not work. But those who do work must have full time jobs since they

earned between 15 to 22 percent of aggregate family earnings going to

93

TABLE 29

The Percentage of Wives in Particular Occu-
pations Within Each Quintile in 1962 and 1971

 

 

 

 

Q‘m‘tile I II III IV v
Occupation 1962 1971 L1962 1971 1962 1971 1962 1971 1962 1971
Professional 2 3 l 1 18 25
Managers 1 1 6 8
Clerical 5 10 2 2 1 14 38 37
Sales 1 2 1 2 6 4 4 4
Craftsmen [ 7 1
Operatives 6 8 1 2 8 8 9
Private
Household 3 1 2 2 2 1 1 6
Service 4 5 4 5 4 9 6 3 4
Laborers
Farmers
Farm
Laborers 3 1
Not in the
Paid Labor
Force 90 89 71 63 86 75 77 69 18 8
Total 98 97 94 94 97 96 98 96 96 95

 

 

 

 

 

 

94

that quintile. Although approximately 30 percent of the wives in Quintile
II worked, their contributions were only 14 to 20 percent of the aggre-
gate family earnings going to that quintile. Between 1962 and 1971 the
percentage of wives who worked in Clerical and Kindred workers in this
Quintile doubled. Quintile III had a tripling of the number of wives

who worked in Sales workers and a doubling of the number in Service, ex-
cept Private Household workers. Yet the contributions of these wives were
noticeably lower than those of the first two quintiles. In Quintile IV
the percentage of wives who worked increased over that of Quintile III.
The percentage of wives who worked in Clerical and Kindred workers went
from 1 percent in 1962 to 14 percent in 1971, the largest occupation shift.
The labor force participation of wives in Quintile V was substantial. In
those families in which the wife worked family earnings were greatly
affected. Twenty-five percent of the wives in Quintile V were in Pro-
fessional, Technical, and Kindred workers, while over 37 percent were in
Clerical and Kindred workers in 1971. In short, Table 29 gives some idea
of which wives are working and where.

In summary, although the Gini coefficients seem relatively stable,
the percentage of aggregate family earnings going to each quintile has
noticeably changed. The top two quintiles have gained at the expense of
the lower three quintiles. The percentage of aggregate hours supplied
by the top two quintiles has increased by .6 percent. The percentage
of aggregate hours supplied by husbands has decreased .6 percent. The
deficit was made up by working wives and extended by their increasing
the percentage of aggregate hours which they supplied by 1.2 percent.

In contrast, the percentage of aggregate hours supplied by the bottom

three quintiles decreased 1 percent. Husbands supplied 2.4 percent

95

fewer hours and wives supplied 1.4 percent more of aggregate hours. As

a result relative income dispersion increased slightly.

3.0 The Impact of Related Factors on the Distribution of Family Earnipgs
Having discussed the actual distribution of aggregate family earnings

and the contributions made by working wives in terms of earnings and hours

worked, this section further explores first the underlying distribution

of all husbands'and all wives'earnings, the adjusted Gini coefficients

derived by holding constant the components of Equations (1.2) and (1.3),

and finally, the adjusted Gini coefficients derived by holding the related

factors of Equation (1.5) constant.

3.1 Adjusted Gini Coefficients for Components of Aggregate Family Earnings
The actual Gini coefficients for aggregate family earnings as well as

those for all husbands'and wives' earnings are compared in Table 30. The

TABLE 30

The Actual Gini Coefficients for Aggregate Family
Earnings, AEFAMt; For All Husbands' Earnings, EHt;
and For All Wives' Earnings, swt; From 1962 to 1971

Year AEFAMt §§:_ EW:
1962 .3249 .3036 .3722
1963 .3258 .3056 .3635
1964 .3225 .3029 .3635
1965 .3179 .2916 .3341
1966 .3120 .2890 .3260
1967 .3138 .2961 .3125
1968 .3113 .2978 .3049
1969 .3180 .3023 .3181
1970 .3211 .3046 .3377
1971 .3414 .3277 .3584

Gini coefficients derived from Equation (1.2) show that the distribution
of all husbands' earnings are much more equally distributed than aggregate

family earnings. In contrast, the distribution of all wives' earnings

96

are much more disperse than aggregate family earnings. The addition of
wives' earnings to that of their husbands is such that the distribution
of family earnings is more disperse. These are results that were antici-
pated from the quintile analysis and from the occupational distribution
of wives.

Using Equation (1.3) to partition the annual earnings of husbands
into hourly earnings and hours worked, the earnings of wives into hourly
earnings and hours worked, the components of the two underlying distribu-
tions of aggregate family earnings are examined. First, the marital labor
force participation rates are held constant while all else varies. The
adjusted Gini coefficient for this factor held constant is substantially
lower than that of the actual distribution of family earnings as shown in

Table 31. The adjusted Gini coefficients holding the marital labor force

TABLE 31
The Gini Coefficient for Aggregate Family Earnings, AEFAMt, the
Adjusted Gini Coefficients for Constant Marital Labor Force
Participation Rates, vfiNt; for Constant Hours Worked Per Year,
H; and for Constant Hourly Earnings, W; From 1962 to 1971

t --Nt - -
Year AEFAM ij .H W_
1962 .3249 .3249 .3249 .3249
1963 .3258 .2893 .3286 .3215
1964 .3225 .2793 .3266 .3161
1965 .3179 .2656 .3220 .3157
1966 .3120 .3041 .3220 .3050
1967 .3138 .3094 .3285 .3025
1968 .3113 .3022 .3260 .3004
1969 .3180 .2824 .3234 .3045
1970 .3211 .2450 .3243 .3082
1971 .3414 .2389 .3475 .3235

participation rates constant exhibit relatively little cyclical behavior
and continue to decrease. This implies that more married women working
has had a substantial disequalizing impact on the distribution of family

earnings.

97

The adjusted Gini coefficients holding the hours worked by both husbands
and wives constant is consistently greater than the actual Gini coefficients
in Table 31. In Chapter 4, it was found that the relative number of hours
worked between husbands and wives was increasing. Thus, Hz/H; is decreasing
and has had an equalizing impact on the distribution of family earnings.

It was also found in Chapter 4 that the relative hourly earnings between
husbands and wives was increasing slightly. Holding this factor constant
gave lower adjusted Gini coefficients, meaning the changes that have occur-
red between the hourly earnings of all wives and husbands had a disequal-
izing impact on the distribution of aggregate family earnings.

The impact of the hours worked per year by wives and their hourly
earnings can be seen by holding each one of them separately constant. In
Table 32, the adjusted Gini coefficient,holding the hourly earnings of

wives constant, is smaller than the actual Gini coefficients. This implies

TABLE 32

The Gini Coefficient for Aggregate Family Earnings,

AEFAMt; the Adjusted Gini Coefficients for Constant

Hourly Earnings of Wives, W5; and for Constant Hours
worked Per Year by Wives, H5; From 1962 to 1971

 

-t —t
Year AEFAMt Yfl .Hl’.
1962 .3249 .3249 .3249
1963 .3258 .3244 .3265
1964 .3225 .3195 .3235
1965 .3179 .3133 .3185
1966 .3120 .3082 .3134
1967 .3138 .3114 .3163
1968 .3113 .3080 .3170
1969 .3180 .3138 .3202
1970 .3211 .3147 .3215
1971 .3414 .3380 .3431

family earnings.

that the hourly wage earned by wives has had a disequalizing impact on

This may be the result of one in every three wives being

98

in Clerical and Kindred workers, an occupation of relatively low pay which
helps to make the distribution of wives' earnings more disperse. Holding
the hours worked per year constant for all wives resulted in adjusted Gini
coefficients that were larger than the actual Gini coefficients in Table 32,
The implication of which is that hours worked per year had an equalizing
impact.
3.2 Adjusted Gini Coefficients for Related Factors

In order to capture the impacts of the increased relative hourly
earnings between all females and males in the labor force, the decreased
relative number of hours worked between all males and females, the slight
increase in relative hourly earnings between wives and husbands, and the
slight increase in the relative number of hours worked by husbands and
wives, adjusted Gini coefficients were derived holding each related factor
in Equation (1.5) constant. The actual Gini coefficients for aggregate
family earnings are larger than those presented in Table 27 because the
algorithm used for computation deleted more observations than that used

in Equation (1.3). Table 33 has the new actual as well as adjusted Gini

TABLE 33

The Gini Coefficient for Aggregate Family Earnings,
AEFAM ; the Adjusted Gini Coefficients Holding iwfi/iwf,
t t t t t
iHh/in' ij/jwh' and ij/th Constant, From 1962 to 1971

 

t t t t t t t t
Year AEFAMt iwh/iwf iHh/in jww/jwh ij/iHh
1962 .3881 .3881 .3881 .3881 .3881
1963 .3848 .3856 .3851 .3842 .3862
1964 .3797 .3792 .3846 .3783 .3817
1965 .3561 .3933 .3532 .3535 .3577
1966 .3270 .3812 .3458 .3269 .3303
1967 .3316 .3759 .3346 .3331 .3339
1968 .3300 .3735 .3207 .3303 .3352
1969 .3346 .3783 .3350 .3368 .3381
1970 .3384 .3777 .3592 .3400 .3408
1971 .3633 .4004 .3852 .3652 .3660

99

coefficients. In Chapter 3, the relative hourly earnings between all
females and males was found to be slightly increasing over time. Holding
iwi/iwf constant resulted in adjusted Gini coefficients that were larger
than the actual Gini coefficients. Thus the increase in the relative
hourly earnings between all females and males in the labor force has an
equalizing effect on the distribution of family earnings.

Holding the relative number of hours constant while all else varied
resulted in higher adjusted Gini coefficients than the actuals, indicating
an equalizing effect. The equalizing effect of the decrease in iHfi/iH:

is not as great as the decrease in iWi/iwt, since the adjusted Gini coef-

f
ficients of the latter are greater than those of the fermer.

In Chapter 4, the relative hourly earnings between wives and husbands
was fOund to be increasing slightly. Holding jWS/iwfi constant resulted
in smaller adjusted Gini coefficients from 1962 to 1966 and larger adjusted
Gini coefficients after 1966 than were the actuals. The reason may be that
the greatest percentage increases in the relative hourly earnings between
wives and husbands mainly occurred before 1965, with the largest percen-
tage decreases after 1965, as shown in Table 20. Thus increases in the
relative hourly earnings between wives and husbands has a disequalizing
impact on the distribution of family earnings -- and conversely.

Finally, in Chapter 4 the relative number of hours worked by husbands
and wives was found to be increasing slightly. Holding this factor con-
stant resulted in adjusted Gini coefficients that were larger than the
actual Gini coefficients. Therefore the increase in the relative number
of hours worked between husbands and wives had an equalizing impact on the
distribution of family earnings.

In order of importance, the increased relative hourly earnings between

all females and males, WE/Wi, had the greatest equalizing impact on the

100

distribution of family earnings of any of the related factors. The de-
creased relative number of hours supplied by all males to that supplied

by all females, Hi/HE, had a lesser equalizing impact. Of equal magnitude
in equalizing the distribution of family earnings was the decrease in the
relative number of hours supplied between wives who worked and whose hus—
bands also worked. The increase in the relative hourly earnings between
wives and husbands had a disequalizing impact before 1965. The relative
decreases in hourly earnings between 1966 and later years had an equalizing
impact on the distribution of family earnings.

All of these influences, however, were overshadowed by the increased
marital labor force participation rates of married females. As shown in
Table 28, working wives increased the number of hours supplied to the top
two quintiles more than to the bottom three quintiles, although the proportion
of hours supplied by working wives in the lower quintile has increased
markedly. As a result, and despite the equalizing influences of the other
related factors, the effect of the increased labor force participation of
married women has been to make the distribution of family earnings more

unequal.

CHAPTER 6

A SUMMARY OF FINDINGS AND CONCLUSIONS WITH
SUGGESTIONS FOR POLICY AND FUTURE RESEARCH

1.0 A Summary of Findings and Conclusions

The purpose of this study was to determine how factors related to
the increased labor ferce participation of married women and the increase
itself affected the distribution of aggregate family earnings. The re-
lated factors considered were suggestions and extensions of the existing
literature. They were: ‘(1) the relative hourly earnings between all fe-
males and males, WE/Wfi; (2) the relative number of hours supplied between
all males and females, HE/HE; (3) the marital labor force participation
rates, Ith; (4) the relative hourly earnings between wives and husbands,
jws/iwt; and (5) the relative number of hours supplied between wives and
husbands, jHS/iHfi.

The necessary information to develop a distribution of aggregate
family earnings was obtained from disaggregated occupational data. In
the process several other pieces of information were analyzed. First,
the ratio of aggregate female earnings to male earnings rose from .26 in

1962 to .31 in 1971. During this period the relative number of hours

t

f’ de-

supplied between all males and females in the labor force, Hi/H
creased from 2.47 in 1962 to 2.19 in 1971 at an average annual rate of

2.27 percent. Most of the decrease in the relative number of hours sup-

plied was a direct result of the 2.40 average annual percentage change

101

102

in the relative number of females to males employed in the labor force.
A dampened effect, however, was achieved by more females working fewer
hours per week more weeks per year as compared to males.

Second, the increase in the ratio of aggregate female earnings to
that of males was partially the result of relative hourly earnings between
females and males, WE/Wi, increasing slightly at an average annual rate
of .71 percent. WE/W: was lowest in 1966 with an estimate of .66 and
highest in 1965 with an estimate of .70. In general the relative hourly
earnings between females and males changed very little. Underneath this
stability, however, females in Farmers and Farm Managers increased their
relative hourly earnings position to males at an average annual rate of
9.81 percent. Females in Laborers, except Farm and Mine workers, also
made sizeable gains in relative hourly earnings with a 3.36 percent aver-
age annual percentage change. In Service workers, except Private House-
hold workers, where over 18 percent of the female labor force was located
in 1971, there was a 2.69 percentage annual increase in their relative
hourly earnings to males. In contrast, 33 percent of the female labor
force incurred relative losses in their hourly earnings as compared to
males in Clerical and Kindred workers. This occupation had an average
annual percentage decrease of .89 percent. The relative gains of the
female labor force with respect to the male labor force in hourly earnings,
in the aggregate, outweighed the losses. This pattern may not be perma—
nent, since the average percentage change in the number of hours supplied
by females to that of males is 3.04 percent annually in Clerical and
Kindred workers, and only 1.58 percent annually in Service, except Private
Household workers.

The industrial data lends further caution to the conclusion that

the average annual percentage change in the relative hourly earnings

103

between females and males will remain positive. Relative hourly wage
rates between all females and males decreased at an average annual rate
of .17 percent. Given the large proportion of earnings that is wage or
salary income, the increased hourly earnings between females and males
may not be a continuing phenomenon.

Third, updating the WOmen's Bureau study on the relative annual earn-
ings gap between females and males in the labor force also revealed a
stalmated situation. Using median annual wage or salary income for full-
time, year-round workers, the annual earnings gap between females and
males went from 58 percent in 1962 to 60 percent in 1971. The gap was
closed by at most a 2 percent increase over the 10-year period. In 1956,
the women's Bureau notes that female workers earned 63 percent of what
males earned. The gap has closed very little in 15 years. A similar
situation was fOund in computing the annual earnings gap between all fe-
males and males for all workers although the percentage was much lower.

Thus the annual earnings gap between females and males has decreased
slightly because the relative gains in hourly earnings from non-farm and
farm self-employment income has offset the relative losses in wage or
salary income. This is partially the result of females moving into rela-
tively higher paying occupations such as Managers, Officials and Propri-
etors and Farmers and Farm Managers at an average annual rate of 3.04 and
8.09 percent respectively. In addition, more females are working fewer
hours more weeks per year as compared to males whose occupational struc-
ture has changed little.

Fifth, the information gained with respect to the population of all
husband and wife families was of equal interest. The ratio of aggregate

wives earnings to that of their husbands rose from .17 in 1962 to .21 in

104

1971. The reasons were: (1) the relative number of hours supplied annually
by husbands to that supplied by wives, iHfi/jfiz, decreased from 5.38 in
1962 to 4.47 in 1971; and (2) the relative hourly earnings between wives
and husbands, jWE/iwi' increased eratically from .89 in 1962 to .92 in
1971 at an average annual rate of .41 percent.

The subpopulation of all husband and wife families in which both
worked showed very different results. The ratio of aggregate wives earn-
ings to that of their husbands ranged from .43 to .47 over the period with
an average annual percentage change of .44 percent. The relative number
of hours supplied between husbands and wives in this group increased from
1.97 in 1962 to 2.08 in 1971, While relative hourly earnings between
spouses increased like that for all husband and wife families. For
Hfi/H: to increase while more and more wives are becoming employed, the
number of wives taking part-time jobs has to be markedly increasing. For
the ratio of aggregate hours supplied by all husbands and wives to fall
either more wives are working while their husbands stay at home or more
families have neither employed. In 1971 11 percent of all husband and
wife families fell within these two categories.

The high correlation in hourly earnings between wives and husbands
was found to be stable over time using an analysis of variance model. The
two most important factors in determining the conditional probability that
a wife was in a particular occupation given that of her husband was in a
given occupation were female occupational opportunities and an interac-
tion term between the wife's occupation and that of her husband. It was
found that husbands in higher ranking occupations tended to have wives
in related occupations -— and similarly for husbands in lower ranking occu-
pations -- despite the fact that one out of every three wives was in Cler-

ical and Kindred workers.

105

Finally, using these findings as background and as essential elements
in constructing aggregate family earnings and its distribution, the follow-
ing results evolved. It was found that the actual Gini coefficients be-
tween 1962 and 1971 were relatively constant while displaying cyclical
movements. The Gini coefficients ranged from .3113 to .3414 over the
period. Absolute income dispersion increased by $5,294 increasing rela-
tive income dispersion from 8.93 in 1962 to 8.97 in 1971. Wives in the
top quintiles in 1962 and 1971 supplied approximately 30 percent of the
aggregate hours in that quintile. The slight gains made by all females
in the labor force, however, increased the percentage of aggregate earnings
which they received for that quintile from 27 percent in 1962 to 28 percent
in 1971. Wives who worked in the bottom quintiles substantially increased
the number of hours which they supplied as well as the earnings which they
received.

The impact of the increased labor force participation of married
women was found to make the distribution of family earnings more unequal.
The impact was strong enough to offset the equalizing impacts of the four
related factors. The increased relative hourly earnings between females
and males in the entire labor force had the strongest equalizing impact.
Second in equalizing importance were the decreased relative number of
hours supplied between all males and females and the decreased relative
number of hours supplied between husbands and wives. Although the in-
creased relative hourly earnings between wives and husbands did not con-
sistently have an equalizing impact on the distribution of aggregate family

earnings, in general it did.

106

2.0 Policy Implications

The policy implications to be drawn from the above conclusions are
straight-forward. First and foremost is the enforcement of Equal Pay for
Equal work legislation; particularly in Clerical and Kindred workers. If
the small yet positive average percentage changes in the relative hourly
earnings had the strongest equalizing impact on the distribution of family
earnings, then equalizing the hourly earning inequities between females
and males would aid in equalizing this distribution. Second, day care
centers or something comparable should be established freeing those wives
in lower income families to work if they so choose. The increase in rela-
tive gains in family earnings at the lower extremes of the distribution
as opposed to the upper extremes would decrease relative income dispersion.
Such a plan would also continue to decrease the relative number of hours
supplied between husbands and wives, a factor which had an equalizing im-
pact on the distribution of family earnings also. Finally, since the in-
crease in the relative hourly earnings between wives and husbands had an
equalizing impact on the distribution, more job opportunities as proposed

by Affirmative Action Programs should be encouraged.

3.0 Further Research

 

Originally, the population of all husband and wife families was deci-
ded upon as the core population in order to avoid the heroic assumptions
that would have to be made in order to bridge the gap between a family
earnings distribution and a family income distribution. The extension to
bridge this gap can go in two directions. First, income other than earn-
ings can be incorporated into the distribution. Such income would include

Social Security payments, rents, dividends, interest, pensions, fringe

107

benefits and welfare payments. Second, those families included in the
population could be extended beyond just husband and wife families. An
extension in both directions -- all income and all families would bridge
the gap between this study and an income distribution that really measure
how well off households are. Even without such a complete income distri—
bution, a study of an after tax family earnings distribution would undoubt-

edly be fruitful.

APPEND ICES

APPENDIX A

108

 

 

 

 

APPENDIX A
Occupations Industries

1. Professional 1. Agriculture

2. Managers 2. Construction

3. Clerical 3. Manufacturing

4. Sales 4. Public Utilities
5. Craftsmen 5. Trade

6. Operatives 6. Service

7. Private Household 7. Public Administration
8. Service

9. Laborers
10. Farmers

11. Farm Laborers

Index of Occupations and Industries

chgpations

1. Professional, Technical, and Kindred Workers

2. Managers, Officials, and Proprietors, except Farm Workers
3. Clerical and Kindred workers

4. Sales workers

5. Craftsmen, Foremen, and Kindred Workers

6. Operatives and Kindred WOrkers

7. Private Household WOrkers

8. Service Workers, except Private Household Workers

9. Laborers, except Farm and Mine
10. Farmers and Farm Managers
11. Farm Laborers and Foremen

Industries

1. Agriculture, Forestries and Fisheries, and Mining

2. Construction

3. Manufacturing

4. Transportation, Communication, and other Public Utilities
5. Wholesale and Retail Trade

6. Service and Finance

7. Public Administration

109

 

 

 

 

Om.mv oo.Nm oo.mm Om.Nm om.mm ovoom OH.©m om.mN om.vm ON.N¢ Oh.mm HhmH
Om.mm Oh.mm Oh.Nm ov.Nm Om.mm om.wm OH.hm ov.mN oo.mm Om.Hv ON.om OhmH
Obowm Oh.vm om.mm om.Nm Oh.vN oo.bm om.®m OHoom ov.mm Om.mv ON.®m momH
Om.vm ov.Nm ON.hm oo.mm ON.mN om.mm OH.mm oo.om ON.mm oo.mv ON.©m momH
Om.mm om.Nm Oh.mm om.mm Om.mN Om.©m omohm Om.Hm O®.mm O©.mv ov.®m homH
OH.vm Ob.Hm Oh.mm oo.mm ov.mm Om.hm Oh.hm Om.Hm OH.®m oo.mv oo.bm oomH
Om.mm OH.mm O®.VM oo.mm OH.vN om.hm ON.hm om.Nm OH.®m ov.vv oo.®m momH
oo.mm O®.Hm ON.Nm oo.mm oo.vm O¢.©m om.hm Omomm Om.mm om.mv om.©m vomH
om.Nm omomm ON.Nm ov.mm O®.vm Oh.®m Om.hm ON.Nm ON.om Om.vv ON.hm momH
Oh.Nm OO.Hm ON.mm Om.©m Oh.vm Ob.©m oo.mm ON.Nm ov.©m ON.mv om.©m Nme
mquzmm
oo.Nv oo.vm Om.mm Oh.mm oo.NN ow.Hv om.H¢ oo.Nv OH.mm OH.m¢ oo.Nv HhmH
ON.Nv oo.mm om.mm Oh.mm OH.NN om.Hv OH.Nv om.mw om.mm Om.mv Oh.mv momH
Oh.Hv Oh.mm om.mm om.mm OH.Nm om.Hv Om.N¢ oo.mv Oh.mm oo.mv oo.vv momH
ON.mv Oh.mm om.mm om.mm oo.vm oo.Nv OHomv Om.Nv Ob.mm Om.mv om.mv hmmH
ON.Hv ON.mm Om.mm om.ov Ob.HN om.mv Omomv 00.04 ON.Ov Oh.mv om.vv oomH
Om.Nv Om.mm om.mm OO.Hv Oh.mm omomv Om.Nv ON.H¢ om.ov ON.Om OH.vv momH
oo.Nv om.Nm om.vm om.ov om.mm ON.NV om.H¢ Ov.ov oo.ov om.mv Om.mw vomH
om.H¢ Oh.Nm OH.mm oo.Hv om.vm oo.Nv om.Hv om.ov ov.ov om.mv om.vv momH
ON.N¢ Ob.mm oo.mm omoHv O¢.©N OH.N¢ OB.H¢ Om.ov om.ov O0.0m om.vv Nme
mquz
Sd
P
.m w 3 w ”a a m s. s ,5...
a a. m s we a. m m I... .
M... s s m m w

 

 

 

xmm 024 ZOHfifimDUUO

fl mamde

Hm zommmm mum Mmm3 mmm Mm03 94 mmbom m0 mmmZDZ mwdmm>< mmh

110

 

 

 

 

 

 

no.mN mm.om mH.mN mm.mN hm.mN hm.mm NH.mm Ho.mN 0N.mm mH.mm om.mm HHmH
HN.om oh.mN hh.mN Hm.mN oo.om mH.mm Ho.mm N>.mN hm.Nm 5H.mm mN.vm oan
mm.mN N>.mN Hb.wN Nh.mN hm.mN vm.mm Ho.om mv.mN oo.mm ov.mm mn.mm momH
on.mN mm.om mv.om mm.mN HH.mN hm.vm 5H.om mv.mN 0N.mm vm.mm mm.mm momH
mn.nN hm.om vv.mN Nm.mN mN.mN mm.vm hh.mm om.wN ve.mm 0H.mm vm.mm momH
ov.mN mm.om mH.Hm hw.mN mN.mN on.mm 0H.mm om.mN om.mm mm.mm no.mm oomH
om.mN mv.om mH.mN mn.mN mo.0m Nm.vm mm.mm on.mN mm.mm mH.mm mo.mm momH
vm.mN NH.om vH.mN mm.mN hm.mN Hw.mm mm.¢m ov.mN on.mm vm.mm mN.mm vomH
om.wN ow.mN om.bN mm.mN mo.mN om.mm om.om nN.mN mw.mm No.mm mo.Nm momH
NH.mN vv.mN mn.nN Hm.mN oo.mN NN.mm mm.om om.mN mm.mm oo.>m 0m.mm NomH
mmqmzmh
Hv.mN MH.ov mv.0m mo.mm III Nm.mm om.N¢ mm.mm mN.mm Hm.ov Nm.me Han
hN.mN vm.mm mo.om oo.mm III vm.mm mo.mv oh.mm mo.mm Nm.ov mm.mv oan
hm.mN Nm.mm 0v.0m mh.mm III Nm.ov mm.mv vN.mm Nm.mm mH.hv oo.vv momH
hm.mN ov.H¢ No.Hm mH.vm III MN.Hv vH.¢v on.mm on.mm mm.>v mm.mv momH
0H.mN m>.Hv Hh.Hm mv.vm III mv.Hv mo.vv HN.mm mm.mm mo.hv mn.mv womH
0N.mN om.Hv mm.Hm vm.vm III Nm.H¢ hN.¢v vN.mm mo.ov mv.nv mo.vv oomH
mm.hN HN.Hv Nm.om vm.vm III om.ov mo.vv HH.om NH.ov HN.>¢ mm.mv mmmH
on.hN mH.Hv $0.0m mo.mm III HN.Hv ob.mv mm.mm mH.ov mo.nv nv.vv vomH
ob.nN mo.ov 00.0m OH.mm III vH.Hv mH.mv mm.vm Hm.ov mm.ov nh.mv momH
oo.hN Hm.ov mm.mN hm.mm III HN.ov mN.Nv mN.mm mm.Hv oo.ov NN.vv NomH
mquz
d
mm... a m s an a. e e m a a; 4....
nt. 3 P T. a u 00
cm W M SA 1 I. a 1 9 UI. /
1 a 1 T. a? E a. S I. 6 98
a 1 a o u.3 1 s o a I s so
1 s 1 a o a T. m e 1 _ e O
S S T. A a T. S 1.3
p 3 U Tr3
S on
UL
Han 08 NomH zomm .xmm D2¢ onacmDUUO mm 20mmmm 2mm mamw 2mm amxmoz MMQWB mo mmmZDZ m0¢mm>¢ mme

m mqm¢9

111

 

 

 

 

00.5mm mv.OOOH om.Nmm Nm.hmm m®.®o> mm.mMNH hh.h@NH mm.Nmm Hm.mmHH mN.HH©H hh.NHNH HhmH
hm.m00H mm.NOOH Nm.MBm Nm.on mm.OOh mm.HHNH mm.VNNH mv.vvm Hw.mmHH mo.vmmH mm.mmNH ObmH
QH.mmm NH.Hm0H oo.mmOH Ho.Nvm Hv.mNh Nm.ova no.5vMH Hm.hmm mm.onHH m©.vm©H MH.HNNH mme
0H.Nmm MN.MOOH ho.NMHH HN.Ohm mm.mmr mm.v®NH h¢.thH mm.mmw mm.m©HH mm.mvoH mH.mONH momH
hH.hmm om.mmm Nv.hhm 0H.H¢h wH.H¢h vm.vaH hN.NmMH hv.Hom NN.HONH NO.H©©H mm.HmNH nomH
mm.w®m Nv.mmm mm.hmOH mthOOH mm.¢mo Hm.m®NH mm.hva mo.mom mm.mONH om.NhhH Nm.MNNH oomH
5N.¢¢m HN.mOOH Nv.mOOH hm.SOOH mH.mNh mo.vaH Ho.mMMH Nm.vmm mm.HNNH mN.mm©H m©.OHNH momH
mo.Hmm Nh.Hmm HN.®om mo.HHOH v5.00b mo.ommH mh.HOMH ow.va mm.mONH w©.monH hm.QONH vomH
m©.Nmm 0H.Nmm mo.mmm Nm.NNOH NM.OMh ov.mNNH mv.HOMH HN.OHm Oh.vNNH mv.om@H om.VHNH mme
mm.Nmm Nh.NHm Nm.HNm om.moOH oo.Hmn mH.mHNH no.NmMH mv.th no.0mmH mN.NonH mN.mNNH NomH
mmflfizmh
No.mmHH hm.m®HN 5N.mBOH 5H.omNH III mv.®moH mv.b®hH Nm.omoH mo.ommH mN.hMNN mo.h©mH Han
Ho.mmHH mm.hmom hh.mmOH ON.omNH III Hm.mvoH Hm.mth mN.vN©H om.ONmH mo.O©NN va.mHmH oan
Nm.bmHH vm.mmON mN.mFOH HH.>OMH III vm.mth mh.hva om.mN©H Oh.mNmH hm.MNMN om.mNmH momH
HN.HmHH mo.MNNN om.MHHH ON.mmMH III Hm.mth 00.nomH vm.mmoH oo.mhmH OH.mHMN mm.NmmH momH
MH.thH 0H.¢VNN NN.mHHH 0N.NSMH III MH.ova Hh.mhmH mo.mbmH mH.hmmH mm.HNMN mH.®HmH homH
mm.NONH mm.mNNN hm.mHHH 0H.HHvH III mm.mbhH mm.mmmH hm.m©mH 00.0HOH mm.mmMN om.mva womH
Ho.NhHH hh.omHN om.hmOH vh.mmvH III mw.mvhH Nm.HhmH Nb.hmvH mm.vaH o>.mQMN om.mmmH momH
vo.m0HH mm.vaN 0H.OhOH mm.NmVH III mH.mm5H vm.ommH H®.O¢VH mH.hooH vN.OMMN ov.mNmH vomH
mo.HbHH hm.NHHN mo.¢hOH mm.mmVH III mm.vvnH vo.momH Nm.mNVH No.0moH mB.VQMN mo.mmmH momH
mm.mmHH 0H.omHN Nh.vaH Ho.®th III Hm.Nm®H mm.mth mm.mN¢H mm.vth mo.ommN mm.mmmH NomH
mquz
em a q a an. a. n a. m m m. <3.
mm m W. M an s n a e a u I. /
1 a 1 I. a e e .4 s I. 6 e a d /.

a 1 a 0 Hi. .4. s m m TL... 9 ..

Q S M a m.8 M w I S Upm/

P a u o.
s m .M

 

 

HhmH OB NomH 20mm .xmm 02¢ ZOHfideUOO Mm 20mmmm mmm Mdmw mmm QmMmO3 mmbom m0 mmmEDZ modmm>¢ mme

U mflm¢8

112

 

 

 

 

 

 

wHOH va mmm mvon mHON hmmm mmv mHmN hmmNH homH mhmm HhmH
Ohm mmH vom mmmm ONHN mmmm mow NHmN SHQNH vmmH NHNm OhmH
vmm HmH va Ohvo HMNN mvmm mmm vmmN ONvNH ova OONm momH
QBOH moH mom wvmm momm movm Nmm mmom VOHNH VNVH moms momH
mmOH 00H moN mmhm omvm Nomm mom mmnN ohmHH vHMH HHov bomH
ONOH mMH mmH mmmm vmom Nmmm mom mONN mHQOH mOMH Hhmv oomH
omVH mVH HNH Nmmm hvmm mmom voN omom thm omNH hHHv momH
NmmH me HmH mwom thm thv mHm QNQN mohm HNNH mmmm vomH
mth NmH HmH mmmv mmNm vmmw mmN mmom whom thH mmom momH
NmmH HoH MbH monv Hmmm mvmv mHm mmmN oomm hmHH boom NomH
mmqflzmh
vva vme Nvom mHmv N4 vHVOH MNMHH 00mm vmbm ammo mmNh HbmH
VHvH hhhH momv mmhv mm HmHoH mmVOH NOHm mmmm mmvh HmHh OhmH
vMVH hmmH MHmv oohm no mBBOH mm®OH mmmN ommm mHHb mHNb momH
movH oomH VHmv Hbmm mm VBBOH mNmOH mowm owhm mmho vNOB momH
oHoH mva mVHv vmmm mm mmmOH omOOH MNwN momm meow VHmo homH
m®VH ommH mVHv mHmm mm mBQOH vmmm Homm Hmhm Nmmo wove oomH
omom mHNN vmov ONmm vm mmBOH vam mmHm Nme mmmo @mmm mme
hHmN NhNN Hva mmnm Hm ovmoH mmmm mMHm mmmm HHmo NNmm vomH
NMNN mNNN Hva whom mm mHNOH mem ONom NHmm omoo mth momH
HmNN mva movv 005m on vwmm mvom Noon moon mth vam NomH
mquz
m. m m m. mm .m. m m m m mm new»
mm m u mm. m m... a m w mm.
1 a 1 I. 89 D. 1. S I. 5 93
m M... m a mm m. m m. m I... .m
S S T. A 3 T. S 3.0
p e u as
an

mQdeDOmB ZH .HhmH 08 Nme zomm .xmm 92¢ ZOHB¢m

Q mHmdB

IDUUO mm meoqmzm mmmMmO3 m0 mmmZDz mmfi

113

 

 

 

 

NvmvHOOH mmova oovam vmonooo HmoNNvH mmoenvo momHmm mnHmmVN HHmvmmvH mmmvaN mnNNNmo HemH
HanMOOH mnemoH mmmvmm mommwmo thvmVH hHmmHvo mmmoom mmommvm HHmvmmvH ommnoNN ovmmovo ocmH
oonmbm mmommH vmmmoN omnvmoo HmmmHoH vmmomNn OHova oonova movmmmvH monmomm oommvmo momH
momHoOH mommOH nommmN mNmmmoo vmmvmnH HoomHmo mmmon manomm omOVVHvH mNmnva Hoommmm momH
HvaNoH thmOH omNeON Nmoomom HmommbH mHHeomo Nnmomv ONHmoVN NmmmommH nmNNmNN mONomom somH
mHmHvOH NmmNMH mmvth nmomHom NooNomH OHmmNmo mHhomv oomomvm oonmmNH nvaHmN oHnNmmm oomH
memovH HomomH mHONNH NNmNmmm vHNHNvN HHhvmvo vaNmm onNONmN mmstHNH omomHHN vovvmmv momH
ammoomH HemomH vhNNqH onOHva mmmonm NHONvmm ooOOHv whovaN cmnHomHH oomnwoN mmmmonv oomH
vaOHoH mvvoHH NmanH mvomvom mvobmmm mmmmmom HHmmmm hovmmmm omNmooHH mnmNHHN memosvv momH
vehvvnH nvmovH mmmmmH ommmHmv ONmHth OOVNvmm Nnova mvmvovm mmHMHoHH oNoomom mNomHav NomH
mmH¢zmh
mHoNNnH wNmnmnm momHNvm mommmNo III HmNommbH oomNHoom mmmmhvm memoonm mmovmomH ommvamH HemH
onomoH evomOHm mmonNHm mmmOMHo III HHomHhoH ommmoomH Hmvmmom ommommm NHomNmoH onoommH ocmH
NmmnHmH moomNmm othmmv nvanmv III mmmthmH vamvan Nmmmmmv mmHmmoo smnnmmoH ooHoommH momH
NmomvnH vwmmev mHmmomv mthon III omHOhmmH thmmomH omHmhnv omHnomm HHmvnomH mHmoanH momH
mHmHMON mmbvomv mmomvov omvmmmm III mvmmommH mHmmnmmH cochmnv HNovvoo mnomnva mNmomOmH nomH
mmHnohH ovmovmv omnovoe hoovmmm III NmmnNmmH mvmmmmmH onvhvv HNmmOHo ovmvnomH novwmmNH oomH
Hvaqu HMMNsmv HHNmHHm NNmmHom III mvhmommH onmvhhH mmNomnv momsvoo wmmHonH ovocmVHH momH
mNnHonN mmsvmmv oommnom ommmmmm III Nommmth m>OHNHbH omoommv HNoHonm HNNmova omOOHNHH vomH
mONmHoN sHmmonv oomnmsv mmmmomm III NmthmHH mmmmmooH OhNoome MHannm mmoooomH MHmHoHHH momH
HVNmmoN mmvhon mmmNoov QNNHovm III NhHwommH vaHmmmH Hmeovmv vommomm mmvonomH onHNoOH NomH
mmH¢2
We. a m a. an a. n a. m a an a...
m m a a... a s a n a mm.
a m a m. on m. a s m. m we... a
1 s 1 a o a I. m e I _ e

s s I. A o I s 5.0

p m u 3

as

 

 

m02¢mD029 2H

.HhmH OB NomH 2022 .20HB¢mDUUO 92¢ xmm Mm HHH¢922¢ OHMmoz mmbom mo mmmEDz H¢BOB mm&
m NHm¢B

114

 

 

 

 

hm. mm. mm.~ om.H mm. N>.m no.m mw.a mm.m nm.m no.m Hmmfi
mm. mo.H wa.m no.a hm. mo.m m>.~ mo.H mm.m mo.m m>.v onma
wm. mm.H oo.H mm.a om. mv.~ hm.m mv.a mo.m mm.m mm.v mama
wm. mo.H vb.H hm.H vo. mv.m mm.m vo.a mm.m mm.m om.m moma
mm. om.H oo.a nv.a Hm. 5H.m vm.m om.a mn.m vm.m on.m noma
«M. on. oo.H Hm.a v0. hm.H Hm.m vv.a no.m vo.m nv.m mama
om. III mm.a NH.H mm. mm.H hm.m vv.a oo.N mm.a mo.m moma
Hm. III III ma.a mm. vm.a mH.N mm.H mm.m vn.a om.m voma
Hm. III III va.a om. mm.a NH.N mo.H mv.m mn.a oo.m moma
Hm. III III vo.H vm. Hm.H mm.a mH.H mm.m nm.a mo.m mama
mquzmm
v0. mm.H HB.N mo.m III mo.v mm.v vn.v mm.v om.m No.0 anma
mm. ov.H mw.m mm.m III om.m mo.v om.w nn.v mn.v on.m onma
mm. vv.a Hm.m mw.m III mm.m 0m.v ma.v mv.v mv.v ov.m moma
mu. o~.H mm.m mm.m III mv.m hm.m om.m oa.v No.v om.v mama
we. mH.H wm.m oo.N III 5H.m on.m mm.m mh.m mn.m oo.v wood
vm. hm.a om.~ av.m III vo.m mm.m on.m mm.m vm.m mm.v mood
mm. om.H o~.m ov.m III mm.m mm.m m>.m mm.m ma.m mm.m mama
mv. mm. ha.m vm.m III vm.m 0m.m Ho.m ov.m wo.m mo.v voma
5v. hm. ma.~ m~.~ III o>.~ -.m ow.m Hm.m mm.m om.m moma
Nu. mm. oa.m ma.m III mo.m o~.m mm.m mo.m om.m mm.m mama
mquz
d
mm m m a. mu .m m m. m m m; :2
nt. 8 E .l a u 00
o m m M s A 1 J a 1 e u J
1 a 1 I. a E E 2 s I. 6 E a .d
a 1 a a U.1 1 s 3 a T‘s 9
1 s 1 a o a r. m e 1 _ 3.0
S S T. A 8 TL S T..3
_ D. a u o o
S uh.“

 

 

Ahma OB NmmH 20mm

m mamdfi

~xmm 92¢ ZOHBdeUUO Mm mB<m mUZHzm¢m NAfiDOfi NIB

115

 

 

mm. om. mm. ¢m. hm.o mm. hm. an. mm. mm. mm. Humfi
vm. #0. mm. mm. no.0 mm. mm. or. mm. mm. mm. onma
mm. mm. mo.a mm. NH.H mm. hm. an. om. mm. mm. moma
mm. 00. vo.H vm. mh.o mm. om. on. mm. mm. mm. mama
mm. Hm. ¢m. mm. mo.H hm. om. mm. om. mm. mm. noma
mm. 00. mm. mm. mo.H mm. mm. mu. om. Hm. vm. coma
mm. mm. hm. mm. No.H hm. mm. mm. mm. mm. mm. moma
mm. mm. mm. mm. mo.H ow. Hm. om. om. mm. vm. voma
on. mm. mm. mm. Ho.H hm. mm. mp. om. mm. vm. mama
sh. mm. am. pm. vm.o hm. Hm. om. om. om. mm. moma
1 J J T S H d 0 fig 8 3 w s d
1 H00.”
Wm m m m mm m m m. m. m f.
1 a 1 I. a e e "H s I. 6 e a d
a I a 3 U.3 3 S o a I s E
1 s 1 a o a I. m e 1 _ 3 O
S S T. A a T. S TO
P 9 u 0 O
s u m

 

 

U mam¢9

Ahma OB NomH 20mm .ZOHfideUUO mfldz mmm
BQEB OE m>Hdemm findzmh mum Mmm3 mmm QMMmOS mMDom m0 mmmEDZ moflmm>¢ mmB

116

 

 

NO.H mm. mm. hm. III mm. mm. Vb. mm. mm. mm. Huma
bO.H mm. mm. mm. III mm. vs. vs. mm. mm. mm. Obma
HO.H on. vm. mm. III mm. mm. vb. mm. mm. mm. mama
HO.H mm. mm. mm. Ill mm. mm. Vb. mm. am. on. moma
mm. m5. mm. vm. III mm. om. mu. vm. Hm. 5h. homa
hm. m5. mm. mm. III mm. m5. mu. vm. mm. m5. coma
mo.H vb. vm. hm. Ill vm. Hm. om. vm. Hm. mm. mood
vo.a Mb. mm. mm. III mm. hm. om. vw. mm. mu. voma
mo.H vb. om. mm. III Hm. mm. Hm. vm. Hm. mm. moma
oo.H Nb. mm. Hm. III mm. mm. mm. mm. am. on. mama
3.... m m a my. ® m m m m mm 3.;
Wm m u 2. m w... a m m mm. a
a m a m. f... n a s m. w E e
M... s a a m... M. w m m . mm
p w u E
_

 

 

HbmH OB NmmH 20mm
mfldzmh mmm mmmw mmm QMMMOS mMmm3 ho ammZDZ mommm>¢ mmB

.ZOHafimDUUO Wm quz mmm Hmmfi OB m>Hdemm

m mam¢8

117

 

 

 

vm. mv. om. mp. III mu. «5. mm. mm. mm. mm. Huma
mm. wv. mm. ms. III vs. ms. mm. on. on. mo. onma
mm. ow. mm. mm. III mm. mm. mm. up. on. mo. mama
mm. «m. No.H an. III mu. vb. Hm. vb. an. mo. moma
m5. vv. pm. as. III mp. Nb. vm. mm. mm. «0. mama
om. mv. mm. as. III an. mm. mm. mm. mm. mm. coma
Hm. mv. mm. on. III mu. Hr. mo. mp. mm. mm. mama
mm. «v. mm. as. III as. an. we. mm. mu. mm. vmma
Hm. mw. mm. on. III on. ms. we. mm. mm. mm. moma
mm. mv. mm. as. III Nb. mu. mm. vs. mm. mm. moma
3.... w... m a. mu .m m m. m m E.

n t. a e I a u o o
m.m m u s A 1 I. a 1 e u I. f
1 a 1 t. a e e 1 s r. 6 9 a e
a 1 a a v.1. 3 s a a T.s 1
1 s 1 a o a I. m e 1 _ I.
S S T; A 9 .l S O

n. m u u

  

 

 

 

Han 09 mama 20mm .onammDooo wm mam: mum same 09 m>Ha¢qmm

quzmh mmm “dmfi mam QMMMOZ mmDOm ho mmmZDZ m0¢mm>¢ mm&

H HAM¢B

118

 

 

Hb. HH. bo. Mv.H mm.bv om. vo. bm. Nv.M MN. vb. Hbma
mo. mo. bo. vv.H mb.mm mm. v0. vm. bv.M mm. Mb. Obma
mm. mo. vo. Nb.H OM.MM vm. v0. mm. NH.M om. Nb. moma
Mb. 00. mo. H®.H mM.mM am. we. vm. om.M Hm. mo. mama
mm. mo. mo. om.H mo.mM mm. v0. bm. vo.M am. we. boma
Mb. be. we. Mv.H mb.m¢ Hm. v0. mm. om.m om. ab. coma
Hb. bo. Mo. bM.H mm.mM bv. Mo. vm. bo.N om. Ob. mooa
mm. bo. Mo. QM.H vb.Mm 0v. Mo. vm. Nb.m ma. be. voma
mb. mo. Mo. mM.H Hm.om me. no. bm. 0b.m ma. v0. Mood
om. bo. v0. bN.H OM.m¢ 0v. v0. mm. bm.N ma. mo. Noma
J J
.. .. q a w"... .m m 1.... m m E I...
m m m. M m M. m E T. a u 0 O
T: 8 I p. u I;
1 a 1 t. a e p. .4 s t. 6 e a d
a 1 a o u.3 3 s o a I s e
1 S I 3 0 a I. m p. I _ 1.0
S S T. A a T. s m.w
P % u u m

 

 

Huma 09 mood 20mm .onemmauoo um omaoqmzm mmmxmoz quz

b mqmdfi

OB m>HB¢dmm Mddzmm ho mmmZDz mmB

119

 

 

mm. mo. 00. mo.H Ill mm. Mo. mv. mm.m bH. wv. Hbma
am. we. be. vo.H III mM. Mo. mv. M®.N 0H. bv. Obma
am. v0. v0. vN.H III 0v. Mo. om. mM.m Va. 0?. moma
Hm. Mo. mo. mH.H III bM. Mo. mv. bM.N ma. Mv. mmma
om. No. v0. bo.H III mm. Mo. mm. OM.N ma. vv. boma
mm. Mo. v0. NO.H III 0M. No. mm. OH.N ma. vv. owma
bm. M0. N0. mm. III mM. No. Mm. oo.N vH. vv. moma
be. M0. M0. mm. Ill NM. NO. vm. mo.N va. NV. voma
v0. N0. M0. N0. Ill NM. NO. mm. NO.N wa. 0v. Moma
00. Mo. Mo. om. Ill MM. Mo. mm. om.H Ma. Nv. mwma
1% m a mu. .m m m. m w my. 5...;
mm m M m... m m... M... m m mo
1 a . . I.
a x a m. mu m m s y. 4.. mm m
1 S 1 a 03 T. w p. I . 1.0
s s T. A T. s I.3

D. 8 u 03

S Uh...

 

 

Hnma 09 mama 20mm .onacmouoo um mung: na¢ mo e<ma oe m>Ha<qmm

wmA¢ZMh AA¢ Mm NAAflDZZd QMMMO3 mMDOm m0 mmmEDz A<BOB MIR

2 Hand?

120

 

 

bm. mm. Mm. mm. III b0. Mm. VM. Hb. N0. Vm. HbmH
Vm. mb. mm. mm. III mm. mm. OM. Hb. V0. Mm. Obma
Om. mm. Nb. mm. III b0. mm. mM. mm. we. ob. moma
mV. Hm. Mb. mm. III Ob. mm. NV. Mb. Mo. Hm. moma
mm. QN.H mm. mm. III mm. MO. NV. Vb. No. mb. boma
mm. Hm. MV. Vm. III mo. Hb. mM. 0b. mm. Hm. moma
bm. III om. bV. III b0. Hb. mM. ab. 00. mm. mood
mo. III III Hm. III mo. 00. VM. 0b. mm. mm. Voma
mo. III III Hm. III mm. 00. OM. bb. mm. Hm. Moma
Vb. III III mV. III b0. H0. VM. 0b. om. Vm. Nwoa
mm m m .3... may. .m m m. m w my.
Wm m H mm. m m... a m w mo
1... m a m. m... m. .. .. .. 5 E...
1 S I 8 0 a I. m 3 a T S
s s I A a w. m _
D. W u

 

 

 

Hbma OB NmmH 20mm .ZOHfideUUO wm Mddz mum Fame 08
m>HB¢Amm fiddzmm mum madm mUZHZM<m wqmoom mmB

A mflm¢8

121

 

 

mm. III Ob. mV. III mm. Hm. ON. mm. Om. Mo. HbmH
mm. III b0. mV. III mm. bV. mN. mm. Om. Ho. ObmH
NV. III 00. mV. III mm. Mm. mN. mm. om. mm. ooma
NM. III Hb. mV. III mm. mm. mm. bm. mV. mm. mood
VM. III V0. mV. III Vm. Nm. mN. 00. om. mm. bme
NM. III Hm. VV. III Hm. Om. CM. 00. mV. mm. coma
III III III HV. III MM. Nm. NM. Ho. MV. Ho. mwma
III III III mm. III mm. Om. HM. No. mV. mm. Voma
III III III mM. III Nm. Hm. bN. N0. mV. mm. Moma
III III III HV. III mm. mm. OM. b0. am. 00. Noma
1d 7d T 3 H6 0 a S D W 5.1a HMOW
m... e m. a mu. .m m w... m. w mm
0m w M SA 1 I: 3 1 p. UT: 10
I 1 T. GD. D. 1. S T. 5 98 p.
a 1 8 a U.3 1 S 3 a T.S 1.0
1 s I a 08 T... w P I . TIC
S S T. A .L S
p a H mm
s _

 

 

2 mandfi

HbmH 09 Nme 20mm .MBQDQD Ddfifibm m.zm203 .ZOHBfimDUUO wm quz mmm
BflZB OE m>HB¢Amm ma<zmm mmm HZOUZH mudddm m0 m0¢3 AdfiZZd Z<Hom2 mmB

THE AVERAGE NUMBER OF HOURS AT WORK PER WEEK
PER PERSON BY INDUSTRY AND SEX,

TABLE N

FROM 1962 TO 1971

 

 

U‘

E . s .5

u, H -H H

g a *5 s 8 .
.5 3‘ E’. 8 o '3 8 o 15’. S
o 4.: “4 -.—I .4 a) w-I H I: -r-I
~r-I U) :3 H H 'U > H '14 JJ
~ . = a n a I a 5 a
Year g“ 8 g a. :3 5.. m a. a: u
MALES
1962 47.70 39.20 41.60 42.30 43.00 41.00 42.70
1963 47.60 39.20 42.00 42.30 37.60 41.10 42.70
1964 47.30 38.90 41.70 42.10 42.30 40.50 42.40
1965 48.20 39.90 42.60 43.00 42.50 40.90 42.90
1966 47.80 40.00 43.00 43.10 42.10 40.90 42.70
1967 48.20 39.80 42.80 42.90 42.50 41.50 42.50
1968 47.80 39.50 42.70 43.00 41.90 40.80 42.20
1969 47.70 39.70 42.70 42.80 41.50 41.10 42.00
1970 47.00 38.70 41.60 42.10 40.90 40.50 41.50
1971 47.30 39.00 42.00 42.40 41.10 41.10 41.90
FEMALES

1962 32.90 35.30 37.60 38.00 34.30 33.70 36.70
1963 32.80 34.70 37.60 37.80 34.20 33.60 37.00
1964 33.00 34.90 37.20 37.30 33.90 33.10 36.40
1965 33.30 33.40 37.90 37.50 34.00 33.50 36.90
1966 33.60 34.20 38.30 37.60 33.90 33.50 36.60
1967 33.80 35.30 37.90 36.90 33.30 33.90 36.70
1968 34.30 34.70 37.70 37.20 32.80 33.40 36.10
1969 34.70 34.50 37.80 36.80 32.50 33.60 36.10
1970 34.90 33.20 36.90 35.80 31.70 33.00 35.60
1971 34.40 34.10 37.40 36.10 31.80 34.30 36.10

 

 

THE AVERAGE NUMBER OF WEEKS WORKED PER YEAR

123

TABLE 0

PER PERSON BY INDUSTRY AND SEX, FROM 1962 TO 1971

 

 

: a 2‘

JJ CD 0 -v-I

m H -H H

g 3 u S m

r5 U 0 4U a) I

g H 5 U -H w w c

H s H m o u o 0-H o

0 +1 ‘H 'H-r-l a) 'v-I H C'H
'2 2 a H: E 5 35¢;
y... 2" 8 4 Es I: a: Mr;
MALES
1962 26.15 34.31 42.91 43.54 35.07 34.37 45.10
1963 25.27 35.60 43.17 43.53 34.74 34.43 45.75
1964 26.08 36.49 43.76 44.08 34.54 34.40 45.58
1965 26.21 37.20 44.01 44.22 34.20 34.78 45.85
1966 30.97 38.18 44.66 44.64 34.77 36.57 45.17
1967 30.49 38.70 44.80 44.58 35.00 36.79‘ 45.83
1968 29.64 38.11 44.62 44.77 34.72 36.21 45.46
1969 28.56 37.68 44.20 43.86 34.01 36.81 44.32
1970 28.68 37.11 43.53 43.62 33.98 35.81 44.81
1971 29.45 36.19 43.72 43.12 33.88 35.96 45.18
FEMALES

1962 13.65 26.21 35.26 35.40 25.00 26.85 34.73
1963 12.33 25.47 35.99 35.89 24.70 26.61 35.81
1964 12.81 22.98 36.27 36.40 24.48 26.77 34.57
1965 14.45 19.75 36.39 37.09 24.98 26.97 33.74
1966 18.54 24.42 36.42 36.93 25.31 28.16 32.31
1967 15.85 30.64 36.87 35.67 26.47 28.81 34.15
1968 14.43 31.86 36.48 35.35 25.54 28.61 34.73
1969 16.65 31.66 35.47 35.34 24.87 28.76 34.99
1970 16.33 32.56 35.35 35.05 24.56 28.80 31.87
1971 17.29 34.66 35.90 36.38 25.31 29.20 33.87

 

 

124

TABLE P

THE AVERAGE NUMBER OF HOURS WORKED PER YEAR PER
PERSON BY INDUSTRY AND SEX, FROM 1962 TO 1971

 

 

 

a a 2*

u m o -H

3 s :1 a m

:0 1: o p m m I

: F1 5 o -a m m c

H a n a 3: . a 222

-H m s Hiya 'U a rA-H u
3‘ a g '3 3 3 m 'Svfi 8
Year :6: 8 0423 B m (1.44.)
MALES
1962 1247.25 1345.15 1785.15 1841.83 1508.00 1409.32 1925.94
1963 1202.83 1395.50 1813.19 1841.41 1306.21 1415.26 1953.37
1964 ,1233.48 1419.33 1825.00 1855.90 1461.16 1393.39 1932.48
1965 1263.26 1484.22 1874.98 1901.54 1453.51 1422.67 1966.86
1966 1480.21 1527.09 1920.59 1924.07 1463.97 1495.52 1928.72
1967 1469.80 1540.29 1917.60 1912.47 1487.31 1526.95 1947.84
1968 1416.93 1505.30 1905.30 1924.95 1454.68 1477.77 1918.46
1969 1362.53 1495.84 1887.50 1877.27 1411.29 1512.94 1861.54
1970 1348.03 1436.26 1810.69 1836.41 1389.70 1450.30 1859.55
1971 1392.88 1411.23 1836.36 1828.30 1392.58 1477.78 1892.96
FEMALES

1962 488.97 925.26 1325.84 1345.31 857.49 904.78 1274.67
1963 404.33 883.85 1353.10 1356.59 844.73 894.25 1325.09
1964 422.65 801.99 1349.13 1357.61 829.87 886.05 1258.37
1965 481.24 659.66 1379.13 1390.76 849.29 903.46 1245.07
1966 622.85 835.24 1394.73 1388.39 852.05 943.43 1182.61
1967 535.87 1081.73 1397.43 1316.09 881.45 976.54 1253.43
1968 495.10 1105.38 1375.32 1315.17 837.56 955.71 1253.70
1969 577.73 1092.20 1340.80 1300.68 808.27 966.23 1263.30
1970 569.90 1081.03 1305.34 1254.85 778.48 950.45 1134.57
1971 594.94 1181.98 1342.74 1313.49 804.81 1001.65 1222.56

 

 

TABLE Q

THE NUMBER OF WORKERS EMPLOYED BY INDUSTRY
AND SEX, FROM 1962 TO 1971, IN THOUSANDS

 

 

>1 0‘

H C1 C

4“ CI) 0 -r-'|

U) H 'H H
:3 D «U :5 U) C
vs 4.: 0 u w m 0
C H :3 U 'H Q) -:-l
H :3 H m 0 JJ 0 U I JJ
0 «U ‘H -H 'r-i <1) H ~r-i 5: ‘0
'H U) :5 H H 'U a H H
a s s a '3 a . a 5‘ a
Year :2: u z a. :> t—« m o. n: "4

MALES
1962 2716 4011 13855 3817 7365 7701 2737
1963 2606 4018 14395 4037 7212 7778 2798
1964 2577 4272 14563 3972 7631 8108 2741
1965 2528 4337 15142 3876 7807 8304 2674
1966 2146 4293 15613 3940 8089 8344 2935
1967 2221 4268 15619 4184 8022 8903 2943
1968 2660 4432 16004 4082 8055 8969 3301
1969 2053 4690 16387 4119 8302 9244 3332
1970 2051 4689 15634 4346 8854 9889 3162
1971 2134 5102 15232 4547 9240 10275 3204
FEMALES

1962 838 224 5678 894 6097 12686 1181
1963 803 198 5681 879 6146 13373 1245
1964 821 229 5801 871 6381 13764 1295
1965 781 219 6155 980 6486 14475 1350
1966 635 245 6635 2053 6938 14798 1453
1967 589 252 6913 1143 7284 14972 1566
1968 605 243 6815 1230 7264 16024 1687
1969 523 259 7253 1283 7511 16585 1690
1970 541 281 6906 1294 7928 17054 6906
1971 565 339 6721 1263 8082 17396 1499

 

 

126

TABLE R

THE TOTAL NUMBER OF HOURS WORKED ANNUALLY BY
INDUSTRY AND SEX, FROM 1962 TO 1971, IN THOUSANDS

 

 

 

:1 a 8‘

u m o H

a, H «I H

:3 :3 +3 :3 U)

.U u u u a) I

c '5 3 8 0:3 8 0-8 5

H o 4) tn «4 H a) «a -1 c: H

n 2 a 5: g z g3:
mu 2 a g .5 a a ..s
MALES
1962 3387518 5395376 24733308 7030275 11106448 10853158 5271303
1963 3134569 5607112 26100851 7433760 9420420 11007929 5465529
1964 3178665 6063398 26577459 7371645 11150107 11297582 5296922
1965 3193511 6437048 28390906 7370375 11347557 11813820 5259393
1966 3176527 6555813 29986130 7580842 11842026 11478578 5660801
1967 3264422 6573966 29951057 8001775 11931211 13594186 5732499
1969 2797267 7015484 30930434 7732474 11716505 13982616 6202640
1970 2764816 6734629 28308300 7981056 12304430 14341980 5879886
1971 2972398 7200074 27971468 8313283 12867406 15184201 6065036
FEMALES

1962 376235 207258 7528140 1202708 5228102 1147862 1505391
1963 324673 175003 7686956 1192442 5191703 11958766 1649742
1964 346993 183656 7626284 1182482 5295385 12195590 1629590
1965 375849 144465 8488541 1362949 5508512 13077640 1680642
1966 395512 204634 9254054 1461976 5911535 13960820 1718327
1967 315625 272596 966040 1504289 6420459 14620739 1962874
1968 299537 268608 9372815 1617659 6084034 15314290 2114993
1969 302151 282879 9724811 1668768 6070901 16024902 2134982
1970 308317 303770 9007774 1623778 6171762 16208915 1836873
1971 336140 400692 9024526 1658937 6504471 17424767 1832622

 

 

127

TABLE S

THE HOURLY WAGE RATE BY INDUSTRY

AND SEX, FROM 1962 TO 1971

 

 

0"
E . s .5
8 E4 '3 8 m
'2 3 8 ‘6 .3 8 30 r:
H s H m o u o 0-H 0
U U 'H -H ‘H a) -.-4 ,4 g .H
'3 2 2 g: "a z 1.. a:
Year 2‘ 8 g a. a: a :2: EE :3
MALES
1962 2.05 3.57 3.21 3.11 2.87 3.15 3.02
1963 2.05 3.60 3.25 3.26 3.41 3.26 3.19
1964 2.09 3.69 3.32 3.38 3.10 3.61 3.31
1965 2.34 3.75 3.40 3.48 3.26 3.73 3.42
1966 2.39 3.99 3.49 3.60 3.47 3.74 3.65
1967 2.37 4.11 3.63 3.81 3.66 3.90 3.81
1968 2.65 4.52 3.90 4.00 3.90 4.41 4.11
1969 2.94 5.07 4.27 4.49 4.29 4.62 4.61
1970 3.30 5.54 4.59 4.87 4.66 5.07 5.19
1971 3.11 5.67 4.74 5.19 4.70 5.04 5.43
FEMALES
1962 --- --- 2.31 2.69 2.26 2.29 3.18
1963 --- --- 2.27 2.72 2.33 2.36 3.17
1964 --- --- 2.39 3.01 2.51 2.56 3.45
1965 2.11 4.71 2.35 2.88 2.65 2.66 3.56
1966 1.61 4.74 2.42 2.98 2.64 2.74 3.86
1967 2.23 3.62 2.60 3.37 2.66 2.84 3.80
1968 2.71 3.34 2.82 3.40 2.98 3.16 4.13
1969 2.36 3.85 3.12 3.82 3.08 3.35 4.37
1970 3.73 4.34 3.40 4.22 3.36 3.58 5.13
1971 3.35 4.41 3.42 4.22 3.34 3.87 5.07

 

 

TABLE T

THE AVERAGE NUMBER OF HOURS WORKED PER WEEK
PER FEMALE RELATIVE TO THAT PER MALE BY INDUSTRY, FROM 1962 TO 1971

 

 

 

 

 

 

Us I

Z” a) S 5 3

U y .,.4 )4 .5
U) U 3 U) c
5 u u u m G O
'U .—I :3 0 -.-I In -H
C :3 :4 I6 04-: 0 on
H o u w -H-H m -H -H m
--I U) :1 .—I.-I 'U > an
a s a an a s 813
Year a: u 5: DID E-I m m-H
1962 .69 .90 .90 .90 .80 .82 .86
1963 .69 .89 .90 .89 .91 .82 .87
1964 .70 .90 .89 .89 .80 .82 .86
1965 .69 .84 .89 .87 .80 .82 .86
1966 .70 .86 .89 .87 .81 .82 .86
1967 .70 .89 .89 .86 .78 .82 .86
1968 .72 .88 .88 .87 .78 .82 .86
1969 .73 .87 .89 .86 .78 .82 .86
1970 .74 .86 .89 .85 .78 .81 .86
1971 .73 .87 .89 .85 .77 .83 .86

TABLE U
THE AVERAGE NUMBER OF WEEKS WORKED PER YEAR PER FEMALE
RELATIVE TO THAT PER MALE BY INDUSTRY, FROM 1962 TO 1971
I
8 . 3 ‘3‘ .5
4J . '

g 8 3 8 In '8:
.U IJ o +1 In «:0
.- 2: a :I I. a: s
H o 4.: ‘H -r~l-H a) -.-I was
H m :1 HH '0 > HH
g, 8 5 an 2 B ‘35
year n: u 2: mo c-c m III-r4
1962 .52 .76 .82 .81 .71 .78 .77
1963 .49 .72 .83 .82 .71 .77 .78
1964 .49 .63 .83 .83 .71 .78 .76
1965 .55 .53 .83 .84 .73 .78 .74
1966 .60 .64 .82 .83 .72 .77 .72
1967 .52 .79 .82 .80 .76 .78 .75
1968 .49 .84 .82 .79 .74 .79 .76
1969 .58 .84 .80 .81 .73 .78 .79
1970 .57 .88 .81 .80 .72 .80 .71
1971 .59 .96 .82 .84 .75 .81 .75

 

 

129

TABLE V

THE AVERAGE NUMBER OF HOURS WORKED PER YEAR PER FEMALE
RELATIVE TO THAT PER MALE BY INDUSTRY, FROM 1962 TO 1971

 

 

 

 

 

 

0‘ I

3 <1) 8 .5 F

33 H .H I. g
m s u s m c
p u 0 u w d o
n H s U -H w -H
c S u m o u 0 U u
H U u m -H-H m -H -H m
-H m D H H U > H u
y E» 8 5 8 ‘3 S 8 8 ‘3
ear 6: U z a. D [4 U) D. 'H
1962 .36 .69 .74 .73 .57 .64 .66
1963 .34 .63 .75 .74 .65 .63 .68
1964 .34 .57 .74 .73 .57 .64 .65
1965 .38 .44 .74 .73 .58 .64 .63
1966 .42 .55 .73 .72 .58 .63 .61
1967 .36 .70 .73 .69 .59 .64 .64
1968 .35 .73 .72 .68 .58 .65 .65
1969 .42 .73 .71 .69 .57 .64 .68
1970 .42 .75 .72 .68 .56 .66 .61
1971 .43 .84 .73 .72 .58 .68 .65

TABLE W
THE NUMBER OF FEMALE RELATIVE TO MALE
WORKERS EMPLOYED BY INDUSTRY, FROM 1962 TO 1971

h m I

c:

*3 a .5. '5 "‘
D :1 «OJ 8 In '8 c:
c u o u 0 d o
a H a U -H m H
H :3 m 0 49 U U «U
o u m -H-H m -H -H m
-H m 5 H H v > H H
8 5 g 8 ‘5' fl 3 9 ti
Year ‘2 8 a. D [.I U) DI H
1962 .31 .06 .41 .23 .83 1.65 .43
1963 .31 .05 .39 .22 .85 1.72 .45
1964 .32 .05 .40 .22 .84 1.70 .47
1965 .31 .05 .41 .25 .83 1.74 .50
1966 .30 .06 .43 .27 .86 1.77 .50
1967 .27 .06 .44 .27 .91 1.68 .53
1968 .29 .05 .43 .30 .90 1.79 .51
1969 .25 .06 .44 .31 .90 1.79 .51
1970 .26 .06 .44 .30 .90 1.72 .51
1971 .26 .07 .44 .28 .87 1.69 .47

 

 

130

TABLE X

THE TOTAL NUMBER OF HOURS WORKED ANNUALLY BY ALL FEMALES
RELATIVE TO THAT OF ALL MALES BY INDUSTRY, FROM 1962 TO 1971

 

 

 

 

 

 

 

m I
3 a .5. E 5
g g p g m 5 G
u o u 0 ¢ 0
E '3 a a o '3 8 u '3
H o u m -H-H w -H -H m
.1 ‘2 a g: E a g 33
Year 2‘ 8 g . s I: a: a. .2
1962 .11 .04 .30 .17 .47 1.06 .29
1963 .10 .03 .29 .16 .55 1.09 .30
1964 .11 .03 .29 .16 .47 1.08 .31
1965 .12 .02 .30 .18 .49 1.11 .32
1966 .12 .03 .31 .19 .50 1.12 .30
1967 .10 .04 .32 .19 .54 1.08 .34
1968 .10 .04 .31 .21 .52 1.16 .33
1969 .ll .04 .31 .22 .52 1.15 .34
1970 .11 .05 .32 .20 .50 1.13 .31
1971 .11 .06 .32 .20 .51 1.15 .30
TABLE Y
THE HOURLY WAGE RATE PER FEMALE RELATIVE TO
THAT PER MALE BY INDUSTRY, FROM 1962 TO 1971
m I
E a, s .5 .5
g Q '3 Q m 5 a
g :1 8 4:, .2 a, < .9.
H 8 b «“3 .3 f. a) .2 .9. 13’
H m s H H U > H u
“ “ E ‘3 'L‘ 2 3‘) ‘3 ‘3
Year g‘ 8 z m D a m m-H
1962 —-- ——- .72 .86 .79 .73 1.06
1963 —-- --- .70 .84 .69 .72 .99
1964 -—- --- .72 .89 .81 .71 1.04
1965 .90 1.25 .69 .83 .81 .71 1.04
1966 .67 1.19 .69 .83 .76 .73 1.06
1967 .94 .88 .72 .88 .73 .73 1.00
1968 1.02 .74 .72 .85 .76 .72 1.00
1969 .80 .76 .73 .85 .72 .73 .95
1970 1.13 .78 .74 .87 .72 .71 .99
1971 1.08 .78 .72 .81 .71 .77 .93

 

 

THE MEDIAN ANNUAL WAGE OR SALARY INCOME PER FEMALE RELATIVE

131

TABLE Z

TO THAT PER MALE BY INDUSTRY, WOMEN'S BUREAU UPDATE, FROM 1962 TO 1971

 

 

>" s: 2‘ I':

5 m o -H -H

m H -H H S
5 :s u :1 m s:
’U 4U O «H (D ‘0
C: r—I :1 0 -r4 0) -r1
H :3 H II! 04-! 0 0-9
.3 t3 “3 “:1“: g "‘ :13
I; 5 g as a E ‘31:
Year n: u 21 mo 9 m III-H
1962 --- --- .54 .63 .45 .47 .70
1963 —-- --- .52 .62 .44 .46 .67
1964 --- --- .53 .65 .46 .45 .68
1965 .34 .56 .51 .61 .48 .45 .66
1966 .28 .65 .50 .60 .44 .46 .65
1967 .34 .62 .52 .61 .43 .47 .64
1968 .36 .54 .52 .58 .44 .46 .66
1969 .34 .56 .52 .59 .41 .46 .64
1970 .48 .59 .53 .59 .40 .46 .60
1971 .41 .65 .53 .58 .41 .52 .60

 

 

APPENDIX B

132

APPENDIX B

The Gini coefficient is a measure of concentration that is derived
from a Lorenz Curve. A Lorenz Curve is obtained by plotting the cumula-
tive percent of some aggregate units -- in this case, all husband and
wife families by their annual family earnings level -- on the horizontal
against the cumulative percent of aggregate family earnings accounted for
by these units on the vertical axis. An hypothetical Lorenz Curve is
illustrated in Figure 1. If all families had equal earnings, the Lorenz
Curve would be represented by the diagonal -- the line of equality -- in

Figure 1. Lorenz Curves drawn from actual data, however, will fall below

 

100

Cumulative Percentage
of Aggregate Family

Earnings

 

 

 

 

 

xi xi+l 100

Cumulative Percentage

all Families for a
particular EEEFAMt

FIGURE 1

A Hypothetical Lorenz Curve for Aggregate Family Earnings

 

133

the line of equality. The greater the inequality in the distribution of
aggregate family earnings, the greater the area between the line of equality
and the actual Lorenz Curve. This area is represented by A in Figure 1.

If B is the total area under the line of equality, then the Gini coeffi

cient is defined as the proportion of B found in A; that is, the Gini

 

coefficient = A/B [13, Miller, pp. 220-221]. Algebraically, P
m
. . f . . = _ __ _ __ __ + __
Gini coe fiCient 1 772 (xij+1 xinyij ij+1)'
ij=1
where m is the total number of marital labor force participation pairings,
XI; represents a particular EEEFAMt or family earnings level, and yii re- 9

presents the percentage of aggregate family earnings.

[l]

[2]

[3]

[4]

[5]

[6]

[7]

[8]

[9]

[10]

[ll]

[12]

134

LIST OF REFERENCES

 

 

 

:-
Budd, Edward C. Inequality and Poverty. New York: W. W. Norton
and Company, Inc., 1967.
Cain, Glen. Married Women in the Labor Force. Chicago: University
of Chicago Press, 1966.
Henle, Peter. "Exploring the Distribution of Earned Income." The _
Monthly Labor Review (December, 1972), 16-27. P

 

Metcalf, Charles. "The Size Distribution of Personal Income During
the Business Cycle." American Economic Review (September, 1969),
657-668.

 

Mincer, Jacob. "Labor Force Participation of Married Women: A
Study of Labor Supply," in Aspects of Labor Economics (Washington,
D.C.: National Bureau of Economic Research, 1962), 63-105.

 

"Labor Supply, Family Income, and Consumption."

 

American Economic Review (May, 1960), 574-583.

 

. "The Distribution of Labor Incomes: A Survey with

 

Special Reference to the Human Capital Approach." The Journal of
Economic Literature (March, 1970), 1-26.

 

 

Sorenson, Gary W. Income Changes and Labor Force Participation.
Corvallis: Oregon State University Press, 1971.

 

U.S. Congress. Joint Economic Committee. The American Distribution
of Income: A Structural Problem, by Lester C. Thurow and Robert E. B.

 

 

Lucas. Washington, D.C.: Government Printing Office, 1972.

U.S. Department of Commerce. Bureau of Census. Concepts and Methods
Used in Manpower Statistics from the Current Population Survey. Cur-

 

 

rent Population Report, Series P-23, No. 22. Washington, D.C.:
Government Printing Office, 1967.

U.S. Department of Commerce. Bureau of Census. Income Growth Rates
in 1939 to 1968 for Persons by Occupation and Industry Groups for

 

 

the United States. Current Population Reports, Series P-60, No. 69,

 

Washington, D.C.: Government Printing Office, 1970.

U.S. Department of Commerce. Bureau of Census. Income of Families
and Persons in the United States. Current POpulation Reports,

 

 

135

Series P-60. Washington D.C.: Government Printing Office, various
years.

"Occupation-~Median Wage or Salary Income in 19__ and 1939 of Persons
in the Experienced Civilian Labor Force, By Work Experience and Sex."

1969. No. 75 Table 63
1970 No. 80 Table 60
1971 No. 85 Table 60
"Industry and Race--Median Wage or Salary Income in 19__ and 1939 of :1

Wage or Salary WOrkers 14 Years Old and Over, By WCrk Experience and
Sex."

 

1969 No. 75 Table 62
1970 No. 80 Table 59 ,
1971 No. 85 Table 59 i

"Type of Income in 19__: Families and Unrelated Individuals By Wage
or Salary Income, Non-Farm Self-Employment Income, Fanm Self-Employment
Income, and Income Other Than Earnings, for the United States."

1962 No. 41 Table 13
1963 No. 43 Table 15
1964 No. 47 Table 15
1965 No. 51 Table 15
1966 No. 53 Table 15
1967 No. 59 Table 21
1968 No. 66 Table 32
1969 No. 75 Table 38
1970 No. 80 Table 42
1971 No. 85 Table 42

"Type of Income in 19__: Persons 14 years Old and Over By Wage or
Salary Income, Non-Farm Self-Employment Income, Farm Self-Employment
Income, and Income Other Than Earnings, for the United States."

1962 No. 41 Table 28
1963 No. 43 Table 32
1964 No. 47 Table 32
1965 No. 51 Table 32
1966 No. 53 Table 32
1967 No. 60 Table 16
1968 No. 66 Table 54
1969 No. 75 Table 61
1970 No. 80 Table 63
1971 No. 85 Table 63

U.S. Department of Commerce. Bureau of Census. Income Distribution
in the United States, by Herman P. Miller. Washington, D.C.:
Government Printing Office, 1966.

 

 

[14]

[15]

[16]

[17]

[18]

136

U.S. Department of Commerce. Bureau of Census. The Current Popula-
tion Survey--A Report on Methodology. Technical Paper No. 7.
Washington, D.C.: Government Printing Office, 1963.

 

 

U.S. Department of Labor. Bureau of Labor Statistics. BLS Handbook
of Methods for Surveys and Studies. Bulletin 1711. Washington, D.C.:
Government Printing Office, 1971.

 

U.S. Department of Labor. Bureau of Labor Statistics. Earnings and
Emplgyment and Monthly Report on the Labor Force. Washington, D.C.:
Government Printing Office, various years.

 

 

n
,_.
. t.-

"Persons at Wbrk in Non-Farm Occupations by Full-Time and Part-Time
Status and Sex."

 

1966 VOlume 13 No. 10 Table A-24
1967 Volume 14 No. 3 Table A-24 .4
1968 Volume 14 No. 10 Table A—24 a
1969 Volume 15 No. 10 Table A-26
1970 Volume 16 No. 10 Table A-26
1971 Volume 17 No. 10 Table A—26

U.S. Department of Labor. Bureau of Labor Statistics. Division of
Employment and Unemployment Analysis: Unpublished data supplied by
Curtis Gilroy, 1973.

U.S. Department of Labor. Bureau of Labor Statistics. Marital and
Family Characteristics of WOrkers. Special Labor Force Report.
WaShington, D.C.: Government Printing Office, various years.

 

 

"Occupation Group of Employed Married WOmen, Husband Present, By
Employment Status and Occupation Group of Husband, March, l9__."

1962 No. 26 Table S
1963 No. 40 Table 8
1964 No. 50 Table R
1965 No. 64 Table Q
1966 No. 80 Table Q
1967 No. 94 Table Q
1968 No. 120 Table Q
1969 No. 120 Table Q
1970 No. 130 Table Q
1971 No. 144 Table Q

"Employment Status of Married Women, Husband Present, By Employment
Status and Occupation Group of Husband, March, l9__."

1962 No. 26 Table R
1963 No. 40 Table R
1964 No. 50 Table Q
1965 No. 64 Table P
1966 No. 80 Table P

[18]

[19]

137

(continued)
1967 No. 94 Table P
1968 No. 120 Table P
1969 No. 120 Table P
1970 No. 130 Table P
1971 No. 140 Table P

U.S. Department of Labor. Bureau of Labor Statistics. Work Experience

 

of the Population. Special Labor Force Report. Washington, D.C.:
Government Printing Office, various years.

 

"Industry: Persons With Work Experience in l9__, By Class of WOrker
of Longest Job and Sex."

1962 No. 38 Table A-2
1963 No. 48 Table A-2
1964 No. 62 Table A-2
1965 No. 76 Table A-2
1966 No. 91 Table A-2
1967 No. 107 Table A-2
1968 No. 115 Table A-2
1969 No. 127 Table A-2
1970 No. 141 Table A-2
1971* No. --- Table A-2

"Industry of Workers With Part-Time Jobs: Persons With Work Exper-
ience in l9__, By Class of Workers of Longest Job and Sex."

1962 No. 38 Table A-3
1963 No. 48 Table A-3
1964 No. 62 Table A-3
1965 No. 76 Table A-3
1966 No. 91 Table A-3

"Wage of Salary Workers and Color: Persons with Work Experience,
By Industry of Longest Job in 1967 and Sex."

1967 No. 107 Table A-3

"Industry of Wage or Salary Workers by Color: Percent Distribution
by WOrk Experience on Longest Job in l9__."

1968 No. 115 Table A-3
1969 No. 127 Table A-3
1970 No. 141 Table A-3
1971* No. --- Table A-3

"Occupation: Persons With Work Experience in 19__, By Longest Job
and Sex."

1962 No. 38 Table A-4
1963 No. 48 Table A-4
1964 No. 62 Table A-5

 

138

[19] (continued)

1965 No. 76 Table A-5
1966 No. 91 Table A-5
1967 No. 107 Table A-4
1968 No. 115 Table A-4
1969 No. 127 Table A-4
1970 No. 141 Table A-4

1971* No. --- Table A-4
*1971 is available upon request but has not been officially published.
[20] U.S. Department of Labor. Wage and Labor Standards Administration.

WOmen's Bureau. Handbook of Women Workers. Women's Bureau Bulletin,
No. 294. Washington, D.C.: Government Printing Office, 1969.