Research " THE RATE OF RETURN TO EDUCATION AND THE GENDER EARNINGS DIFFERENTIAL : A COMPARISON OF THE UNITED STATES AND THE REPULIC OF IRELAND " ppt

- -- -- ng sex 65 Private Rates of Return to Educational Attainment by Gender: Ireland 1987 and 1993 Percent...::cccccccssssssecceescccecceccecersesceecessssncenseees 65 Level of Educati

This manuscript has been reproduced from the microfim master UMI films the text directly from the original or copy submitted Thus, some thesis and

dissertation copies are in typewriter face, while others may be from any type of computer printer

The quality of this reproduction is dependent upon the quality of the copy submitted Broken or indistinct print, colored or poor quality illustrations and photographs, print bieedthrough, substandard margins, and improper alignment can adversely affect reproduction

In the unlikely event that the author did not send UMI a complete manuscript

and there are missing pages, these will be noted Also, if unauthorized copyright material had to be removed, a note will indicate the deletion

Oversize materials (e.9., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps

Photographs included in the original manuscript have been reproduced xerographically in this copy Higher quality 6” x 9° black and white photographic prints are available for any photographs or illustrations appearing

in this copy for an additional charge Contact UMI directly to order

Bell & Howell information and Leaming

300 North Zeeb Road, Ann Arbor, MI 48106-1346 USA

800-521-0600 UMI

UNITED STATES AND THE REPUBLIC OF IRELAND

by Maureen Carron Morris submitted to the Faculty of the College of Arts and Sciences

of American University

in Partial Fulfiliment of the Requirements for the Degree

8

UMI

UMI Microform9973425 Copyright 2000 by Bell & Howell Information and Learning Company

All rights reserved This microform edition is protected against unauthorized copying under Title 17, United States Code

Bell & Howell Information and Learning Company

300 North Zeeb Road P.O Box 1346 Ann Arbor, Mi 48106-1346

supported me through the ups and downs of this entire process First, my parents,

Desmond and Carolyn Carron, whose endless encouragement in all my academic pursuits has inspired the completion of this degree Second, my husband, William J Morris, III, who provided constant support and reassurance that someday this paper would be

finished! And third, my father-in-law, William J Morris, Jr., who received his Doctor of

Philosophy in Economics from American University in 1968, whose persistent reminders

to continue on despite the many distractions of life were essential in bringing this

dissertation to completion Most importantly, this dissertation could not have been

written without the countless hours he and my mother-in-law, Veronica Morris, spent taking care of my daughter Brigid, and then, her younger brother, William (Liam), so I could type away at the computer To you goes my most heartfelt and sincere thanks for all the time, effort, and love you put into making this degree a reality

UNITED STATES AND THE REPUBLIC OF IRELAND

BY Maureen Carron Morris ABSTRACT International evidence shows that while the size of the gender earnings differential

varies across countries, the U.S and Ireland seem to have some of the largest differentials

among western European and Scandinavian countries There are several possible factors including differences in the prices of labor market skills, differences in the quantities of

labor market skills, and differences in unobservable factors between men and women in

the U.S and Ireland that prevent the reduction of the gender earnings differentials in the two countries The aim of this dissertation is to measure and analyze the separate

components of the gender earnings differentials including a decomposition of the residual for both the U.S and Ireland to determine what is impeding wage convergence within each country Three models are used: the first estimates the rate of return to education; the second determines the gender earnings differential; and the third decomposes the gender earnings differential residual into predicted and residual components using a decomposition of the cross-country gender earnings differential for the U.S and Ireland The findings indicate that both the U.S and Ireland experienced an increase in the rate of return to education for women which has been shown to have a significant impact in narrowing the gender earnings differential Additionally, the results indicate that the

ii

combined with an increase in the mean female percentile in the male wage distribution caused a substantial narrowing of the gender earnings differential between 1987 and

1993 However, in Ireland, gender-specific factors play an important role in the gender earnings differential, and, as a result, even though the productivity differential decreased along with an increase in the mean female percentile in the male wage distribution, the discriminatory differential increased which prevented a significant narrowing of the gender earnings differential between 1987 and 1993

lil

hours reading and editing numerous drafts of this paper Julia always provided rapid feedback, often sacrificing time on the weekend to read my work, for which I thank her Julia was a source of constant encouragement and support throughout the entire process While she provided the necessary guidance, she allowed me to learn for myself the

aggravations and joys of research, for which I am grateful Additional thanks are due especially to Tom Husted who provided much assistance and guidance in the early stages

of this dissertation Tom read and revised many drafts of the original proposal for this work And finally, thanks are due to Bob Lerman who provided insightful comments that helped bring the paper to completion

iv

The Gender Earnings Differential The Earnings Differential Residual Decomposition

CO er mem er cee e eects eee meee e eee ee eee eee Hee tessa wee eeeeeeeee “““e*.eeseeeeeee

ĐC Hi HẢO G4 0 0 0 00000 00400000 00100 0000010000040 09000 0 0 0 400 0 00 00 410001006 i40 000 0/0 04004 tên bê nh nh“ H06 1e ng 6606 mene

ee ee rere rere rrr rrr rrr Tritt

18 22

l0

Strengths and Weaknesses of the Data - - se

l0 1a —

H01)

Conformability of the Data - - - Ăn mem

Conformability of the Irish data across the two years Conformability of the U.S data across the two years Conformability of the data across the U.S and Ireland

Model 2: The Gender Earnings Differential Decomposition

Research Questions .ccccccccssceseecececeeeeeceseereroeees

VI

VIL

Research QQuestions - mm mm 25

Frequency of Marital Stafus -.- - «<< se

U.S Eamings Ratios by Marital Status

Irish Earnings Ratios by Manital Status

Earnings of U.S and Irish Women by Marital Status

The Link Between Education, Marital Status, and Earnings

RESULTS OF MODEL 2: WAGE DIFFERENTIAL

VIII RESULTS OF MODEL

<F.F.Dc 9Ó GÓP G9 HH G6 0 G0 00 9 G0 6 60 000 0 00 000000000000 00 0 00 e6

3: WAGE DIFFERENTIAL RESIDUAL DECOMPOSITION À - - S<<<< <<

The Race Variable

The Female Variable

Cr rr rrr

CO ere rer ecccccaceserecnoensecasecseneeeseseseoseseosace -

The Marital Status Variablc g4 5 ca se .c se The Work Status Variable P000 0”966000600000006060000006009000990080601®0e09008660 064966

APPENDIX B: Test of the Experience Proxy Variabile -

APPENDIX C: The Selectivity Bias Problem and Methods of Correction

Heckman Probit Equation - - sscss 4 1 eceesee Extended Eamings Function ẶẶScccSĂ se neeeeeres APPENDIX D: Regression Results from Model 2 Ặ -ĐQ- 2<<<<5

APPENDIX E: Regression Results from Model 3 .-

APPENDIX F: Results and Analysis for Part-Time Workers

Average Hourly Earnings and the Gender Earnings Ratios

Frequency of Marital Status for Part-Time Workers

U.S Earnings Ratios for Part-Time Work by Marital Status

Irish Earnings Ratios for Part-Time Work by Marital Status

U.S and Irish Part-Time Earnings by Marital Status

APPENDIX G: Average Hourly Earnings by Marital Status

t)

10

Il

12

13

14

Reorganization of the Race Variable, U.S Data: 1987 and 1993 36

Reorganization of the Education Variable, U.S Data: 1993 36

Reorganization of the Education Variable, Irish Data 38

Original and Final Industry Categories, U.S and Ireland: 1987 Data 40

Original and Final Industry Categories, U.S and Ireland: 1993 Data 4I Original and Final Occupation Categories, U.S and Ireland: 1987 Data 43

Original and Final Occupation Categories, U.S and Ireland: 1993 Data 44

Reorganization of Marital Status Variable, U.S and Ireland 46

Rates of Return to Education: United States and [reland, 1987 and 1993 (Percent) 4dd43a15ăĂ H,.),))L), 62

Female Returns to Educational Attainment: U.S and Ireland, 1987 x- ` 17 — 64

Female Returns to Educational Attainment: U.S and Ireland, 1993 (PETC) 64

Private Rates of Return to Educational Attainment by Gender: United States, 19§7 and 1993 (PercerI) - ng sex 65 Private Rates of Return to Educational Attainment by Gender: Ireland 1987 and 1993 (Percent) ::cccccccssssssecceescccecceccecersesceecessssncenseees 65 Level of Education Completed by Gender: United States, 1987 and 1993 (20 d3 67

Completed and by Age Cohort: Ireland, 19§7 - -=<s ===s=

Average Hourly Earnings for Full-Time Work by Level of Education

Completed and by Age Cohort: United States, 1987

Average Hourly Earnings for Full-Time Work by Level of Education

Completed and by Age Cohort: Ireland, 1993 ccccccscccssseeeeeeeee

Average Hourly Earnings for Full-Time Work by Level of Education

Completed and by Age Cohort: United States, 1993

Frequency of Marital Status of Full-Time Workers by Gender:

United States and Ireland, 1987 (Percent in sample)

Frequency of Marital Status of Full-Time Workers by Gender:

United States and Ireland, 1993 (Percent in sample) - -

Average Hourly Earnings for Full-Time Work by Marital Status:

United States, I9§7 and 1993 cu xxx ce

Average Age of Full-Time Workers by Marital Status: United States,

U.S and Irish Women, 1987 and 1994 cm se se se eeeeeze

Frequency of Union Membership for Full-Time Workers:

United States and Ireland, 1987 (Percent in sample) Frequency of Union Membership for Full-Time Workers:

United States and Ireland, ¡993 (Percent in sample) -

xi

74

75

80

Average Hourly Earnings for Full-Time Work by Union Membership:

United States, 1992 - - - - Q-QQQQ HH HH nu HH KH nen n1 8l

Average Hourly Eamings for Full-Time Work by Union Membership:

Decomposition of the Gender Wage Differential, Ireland: 1987 and 1993 89

Decomposition of the Gender Wage Differential, Gross (Unadjusted)

Wages: United States, 1987 and 1994 - «sen H9 nen mm 90

Decomposition of the Gender Wage Differential, Gross (Unadjusted)

Wages: Ireland, 19§7 and 1992 - sen n1 ng mm 92 Gender Earnings DifferentiaÌS - - 5 x3 ng gu 100 Mean Female Percentile in the Male Wage Distribution - 100

Log Wage Differential Decomposition, United States and Ireland:

L987 and 1993.00 103 Residual Decomposition Effects: 1987 and 1993 Q2 Sex 110

Mean Female Percentile in the Male Wage Distribution,

United States and Ireland: 19§7 and 1993 - QG Sex eee 119

Percentage of Females Earning Less than Males,

United States and Ireland: 1987 and 1993 (by percent) - 120 Composition of Sample by Race: United States, 1987 and 1993 (Percent) 126

xii

Full-Time and Part-Time Workers by Marital Status:

United States, 1987 and 1993 (Percent ¡in Sample) Full-Time and Part-Time Workers by Marital Status:

Ireland, 1987 and 1993 (Percent in Sample) - - << ses<se=szs

Frequency of Industry by Gender: United States and Ireland,

1987 and 1993 (Percent in Samp̀) - - - - 5 - S c nen 518525 ce

Frequency of Industry by Work Status: Women in the United States and Ireland, 1987 and 1993 (Percent in Sample) - «<< sex

Frequency of Occupation by Gender: United States and Ireland,

1987 and 1993 (Percent in SampÌe) - - - sec Frequency of Occupation by Work Status: Women in the United States

and Ireland, 1987 and 1993 (Percent ¡in Sample) <-

Regression Output using EXPPROX and YREMP for 1987 Irish Data Regression Output using EXPPROX and YREMP for 1993 Irish Data Female Labor Force Participation Rates: United States and Ireland

Selectivity Corrected and Uncorrected Rates of Return to Education: United States, 1987 (percent) - - Ác HH HH ng re Heckman Coefficients and Standard ETTOrFS - 5S se se ezsmersz Regression Coefficients and Standard Errors - «<< <c< «<< Coefficients from Regression Analysis of lusdwage -

Xili

62

63

64

65

66

67

68

69

70

71

73

74

75

76

77

78

Coefficients from Regression Analysis of lusdwage - 154

Coefficients fom Regression Analysis of Iwkwage - 155

Coefficients fom Regression Analysis of Iwkwage . -~ - 156

Coefficients from Regression Analysis of Ïrwage -~ 157

Coefficients from Regression Analysis of Ìrwage - 158

Coefficients from Regression Analysis of lInusdwg - - - 159

Coefficients from Regression Analysis of lnusdwg - 160

Coefficients fom Regression Analysis of Irwkwage - 161

Coefficients fom Regression Analysis of Irwkwage - 162

Coefficients from Regression Analysis of lnusdwg - - 163

Coefficients om Regression Analysis of lrusdwg - -~- 164

Coefficients from Regression Analysis of Irwkwage -«- 165

Coefficients from Regression Analysis of Irwkwage -. -«- 166

Average Hourly Earnings for Part-Time Work by Level of Education Completed and by Age Cohort: Ireland, 1987 -. - 168

Average Hourly Earnings for Part-Time Work by Level of Education Completed and by Age Cohort: United States, 1987 169

Average Hourly Earnings for Part-Time Work by Level of Education Completed and by Age Cohort: Ireland, 1993 - 170

Average Hourly Earnings for Part-Time Work by Level of Education Completed and by Age Cohort: United States, 1993 171

XỈV

United States and Ireland, 1993 (Percent in sample) - -

Average Hourly Earnings for Part-Time Work by Marital Status:

United States, 1987 and 1993 - nong nen

Average Hourly Earnings for Part-Time Work by Marital Status:

Ireland, 1987 and 1993.00

Average Hourly Earnings for Part-Time Work by Marital Status:

U.S and Irish Women, 1987 and 1993 :cccsseesseescsseteeeseeenseeeeeees Average Hourly Earnings for Full-Time Work for Married Individuals

by Level of Educational Attainment and Age Cohort: Ireland, 1987

Average Hourly Earnings for Full-Time Work for Separated, Divorced,

or Widowed Individuals by Level of Educational Attainment and

Age Cohort: Ireland, I9§7 «7 sec nen

Average Hourly Earnings for Full-Time Work for Single Individuals

by Level of Educational Attainment and Age Cohort: Ireland, 1987

Average Hourly Earnings for Full-Time Work for Married Individuals

by Level of Educational Attainment and Age Cohort: U.S., 1987

Average Hourly Eamings for Full-Time Work for Separated, Divorced,

or Widowed Individuals by Level of Educational Attainment and

Age Cohort: U.S., I9§7 -Ă SH HH HH 038 56

Average Hourly Earnings for Full-Time Work for Single Individuals

by Level of Educational Attainment and Age Cohort: U.S., 1987 Average Hourly Earnings for Full-Time Work for Married Individuals

by Level of Educational Attainment and Age Cohort: Ireland, 1993

Average Hourly Earmings for Full-Time Work for Separated, Divorced,

or Widowed Individuals by Level of Educational Attainment and

Age Cohort: Ireland, I994 - + vn ng mm

93 Average Hourly Earnings for Full-Time Work for Married Individuals

by Level of Educational Attainment and Age Cohort: U.S., 1993 185

94 Average Hourly Earnings for Full-Time Work for Separated, Divorced,

or Widowed Individuals by Level of Educational Attainment and

Age Cohort: U.S., I994 - - HH HH gen 186

95 Average Hourly Earnings for Full-Time Work for Single Individuals

by Level of Educational Attainment and Age Cohort: U.S., 1993 187

xvi

l

2

GDP and the Wage Differential: Ireland,

Percent Change Over Previous Year 1987 and 1993

GDP and the Wage Differential: U.S.,

Percent Change Over Previous Year 1987 and 1993

XVil

ww een wccccaresccsereveseccsces

Ôi

In the 1980s, the United States and the Republic of Ireland experienced similar gender earnings differentials at 67 percent and 68 percent, respectively International evidence shows that while the size of the gender earnings differential varies across countries, the U.S and Ireland seem to have some of the largest differentials among western European and Scandinavian countries There are several possible factors that may explain why the U.S and Ireland have such relatively large gender earnings

differentials One explanation focuses on prices It could be the case that the prices of the labor market skills are different in the U.S and Ireland which would account for the larger earnings differentials A second explanation focuses on the quantities of labor market skills It could be the case that men and women in the U.S and Ireland have different human capital endowments which could be preventing a reduction in the magnitude of the gender earnings differential The differences in the female labor force participation rates provide support for this explanation The participation rate for Irish women is substantially lower than the corresponding level for American women, which indicates that perhaps barriers exist that prevent the entry of Irish women into the labor force These potential barriers may be preventing a reduction in the gender earnings differential A final explanation focuses on unobservable factors It could be the case that unobservable factors exist in the U.S and Irish labor markets such as unobservable gender-related productivity differences between men and women, or unobservable labor market wage-setting institutions, that prevent a reduction in the magnitude of the gender earnings differentials in the U.S and Ireland

1

determine what it is that is impeding wage convergence for each country Other

interesting analysis that falls under this broad topic includes a calculation of the rate of return to educational attainment for the U.S and Ireland with particular attention paid to the differences in the returns to education for men and women The rate of return to education provides an indicator of the importance of education on earnings for men and women These earnings, or rewards to human capital endowments, are different for men and women, which gives rise to the gender earnings differential However, the

differences in the rewards to productivity-related attributes of men and women comprise only one aspect of the gender earnings differential

One goal of this paper is to decompose the gender earnings differentials for the U.S and Ireland in order to determine the magnitude of each of the components of the differential, which will indicate the portion of the pay gaps in the U.S and Ireland that result from differences in the rewards to human capital endowments of men and women (the productivity differential) and from gender discrimination (the discriminatory

differential) This paper will also address the issue of the unobservable residual of the gender earnings differential Specifically, to what extent are other factors that are hidden

in the residual, such as differences in gender-specific productivity-related characteristics, wage-setting institutions, and gender discrimination, responsible for the gender earnings differential? What is the impact of gender-specific factors on the male-female wage differential? Finally, what is the impact of labor market factors (i.e., labor unions,

collective bargaining, and other wage-setting labor market institutions) on the

male-female wage differential? As previously mentioned, the underlying focus will be to determine what is impeding the convergence of male and female wages in the U.S and

Ireland

Republic of Ireland, referred to throughout as Ireland Two years of data will be

compared: 1987 and 1993 The 1987 data will provide a consistent estimate for

comparison to other work in the literature that was conducted using data from the U.S that ranges in the 1980s Additionally, the 1993 data will provide a relatively recent update to the 1987 estimates in order to illustrate any trends that may become apparent

The rate of return to educational attainment for both men and women in the U.S and Ireland will be estimated using Mincer’s human capital earnings function and an extended earnings function in order to (1) determine the importance of education on earnings and (2) assess the impact that gender differences in educational attainment have

on the gender earnings differential in the U.S and Ireland The gender earnings

differential will be estimated for the U.S and Ireland using the Oaxaca-Blinder

decomposition model in order to determine how much of the gender wage differential is due to differences in the human capital characteristics of men and women, such as

educational attainment, and how much of the gender earnings differential is due to other unobservable determinants, such as gender discrimination Finally, the significance of these other unobservable determinants of the gender wage differential, such as

gender-specific factors, wage inequality, and the wage structure within each labor market, will be estimated for the U.S and Ireland using a decomposition of the residual from the gender wage differential

There are several original aspects of the work performed in this dissertation, including a calculation of the rate of return to education for Ireland Additionally, while the gender earnings differential has been calculated for Ireland for 1987 (Callan, 1991), it has not been computed for 1993 A cross-country comparison has not been made

between the U.S and Ireland with respect to the gender wage differential residual

decomposition In addition, while Blau and Kahn have decomposed the wage differential

these gaps

O «oes { the Stud This chapter has provided a brief overview of the work to be done in this paper Chapter two presents and discusses the possible factors that may impact the magnitude of the gender earnings differentials in the United States and Ireland Chapter three presents

a review of the literature that is relevant to the work contained in this paper as well as a discussion of the original contribution of this dissertation Chapter four discusses the data and some of the problems associated with the data that will be used in this paper Chapter five presents the three models that will be used to estimate the rate of return to investment in education, the gender earnings differential, and the decomposition of the wage gap residual Chapters six, seven, and eight present the results of the estimations for each of the three models, respectively And chapter nine links the results together and discusses the broad findings of this dissertation in the concluding discussion

This chapter is organized as follows: first, the different factors present in the

United States and Ireland that may impact the gender earnings differentials in the two countries are identified Second, each possible factor is discussed in turn with particular attention paid to how that factor either facilitates or impedes the reduction of the gender earnings differential in each country Specifically, the differences in the labor market institutions between the United States and Ireland are discussed; the differences in the macroeconomic conditions present in the U.S and Ireland during the late 1980s and early 1990s are discussed; and the differences in the human capital characteristics of women in the U.S and Ireland are discussed, focusing on the differences in the labor force

participation, educational attainment, and marital status of women in the two countries Finally, a concluding summary brings together the effects that each of the various factors have on the gender earnings differentials in the U.S and Ireland

Motivati The United States and Ireland experienced similarly large gender earnings

differentials in 1987 at 69 percent and 68 percent, respectively There are several factors that affect the magnitude of the gender earnings differentials in the two countries

including differences in the labor market institutions, differences in the macroeconomic conditions, and differences in the human capital characteristics of women Each of these differences will be discussed in turn in the following sections

result from differences in the labor market wage-setting institutions present in the two countries The following sections discuss the differences and similarities between the U.S and Ireland regarding gender-specific legislation and wage-setting institutions that either facilitate or impede reductions in the magnitude of the gender earnings

differentials

Gender-specific Legislation:

The United States and Ireland have both implemented legislation designed to reduce discrimination against women in the workplace and increase equality of pay and opportunity However, despite the implementation of anti-discrimination legislation and equal pay policies, gender earnings differentials continue to persist in both the U.S and Ireland The following discussion considers why the gender earnings differentials have not narrowed

The United States began implementing equal employment opportunity and

anti-discrimination policies in the 1960s while Ireland did not begin such efforts until

1974, and only then as a result of an EC Directive However, the implementation of

equal pay legislation has had a relatively small effect on the gender earnings differential mainly because women tend to be “considerably segregated” by occupation and industry

in both the U.S and Ireland (Blau and Kahn, 1995, 111) Occupational segregation contributes to the magnitude of the gender earnings differential because women

employed in female-dominated occupations will be affected minimally by equal pay policies that require equal pay for equal work within the same occupation Occupational segregation is prevalent in both the U.S and Ireland which impedes the reduction of the gender earnings differentials in the two countries.

States and Ireland have legislation regarding family leave policies; however, the effect of such policies generally is to widen the gender earnings differential because interruptions

in a woman’s career tend to lower the average wage Additionally, the possibility arises that employers will discriminate against women due to the costs associated with

mandated leave policies

The intent of the equal pay, anti-discrimination, and family leave policies in both the U.S and Ireland is to attempt to remedy the existing unequal labor market outcomes for men and women; however, legislation may not be the only solution If the unequal outcomes are due to unequal opportunities in the labor market, then legislation will be effective in narrowing the gap between male and female earnings by reducing and/or preventing wage discrimination However, if the unequal outcomes are due to unequal incentives to participate in the labor market or due to differences in tastes and

preferences, then policies will be ineffective in narrowing the gender earnings

differential

Wage-Setting Institutions:

It has been determined that a large portion of the gender earnings differential in the U.S stems from “interindustry or interfirm wage differentials” that arise because the U.S does not have a system of centrally determined pay (Blau and Kahn, 1995, 111) In contrast, Ireland has a somewhat more centralized system of wage determination which results from the strong collective bargaining process in Ireland Systems of

wage-determination that are less centralized tend to have wider gender earnings

differentials because there is much more variation in wages across firms and industries The United States has a decentralized system of wage-determination which works to

Additionally, the collective bargaining process in the United States is

decentralized with only about 20 percent of the men and women who work full-time

belonging to unions (Blau and Kahn, 1995, 115) However, the role of unions and

collective bargaining in Ireland is significant; in fact, evidence has shown that unions play an important role in reducing the magnitude of the gender earnings differential in Ireland (Reilly, 1995, 221)

Labor market wage-setting institutions, which include the system of

wage-determination, unions, and the collective bargaining process, present in the U.S tend to impede reductions in the U.S gender earnings differential; whereas, these same labor market wage-setting institutions present in Ireland tend to facilitate reductions in the Irish gender earnings differential

Macroeconomic Conditions The gender earnings differentials that persist in the United States and Ireland may result from differences in the macroeconomic conditions that prevailed in the late 1980s and early 1990s in the two countries The following section discusses the differences and similarities between the U.S and Ireland regarding the GDP growth rate and the

unemployment rate and their effect on the gender wage differential

The graphs below illustrate for the U.S and Ireland how the growth of GDP compares to the real average hourly wage differential for the years 1987 through 1993 Both GDP and the gender wage differential are measured as a percent change over the previous year The top curve in the Irish graph represents the percent change over the previous years’ growth in GDP and the lower curve represents the percent change in the real average hourly wage differential In Ireland, the growth of GDP is increasing over

differential and the lower curve represents the percent change in GDP growth Like Ireland, the U.S GDP growth rate is increasing despite a slight drop between 1990 and

1991 The wage differential appears to decrease significantly between 1987 and 1989 and then move along with the growth rate

The main result is that for both countries, there does not seem to be a clear

connection between the change in GDP and the change in the gender wage differentials The reason is that when the growth rate increases, wages tend to increase, but if average female wages rise more than average male wages there will be a narrowing of the gender earnings differential which occurred for several years in both countries However, if average male wages rise more than average female wages there will be a widening of the gender earnings differential which also occurred for several years in both countries The net effect is that, in general, the growth rate of GDP increased and the gender wage

differential narrowed over the time period of 1987 to 1993

Both the United States and Ireland experienced tight labor markets during the late 1980s and early 1990s The unemployment rate in the U.S and Ireland has been steadily decreasing over the last two decades, which indicates that more people are entering the labor force The fact that the female labor force participation rates for both the U.S and Ireland increased between 1987 and 1993 indicates that more women are entering the labor force When increasing numbers of women enter the labor force the tendency is for the gender earnings differential to widen because more and more women are working and those who are entering the labor force later are likely to have less education and earn lower wages than those already working, which pulls down the average female wages and widens the gender earnings differential The role of female labor force participation and women’s educational attainment on the magnitude of the gender earnings differential will

be more fully developed in the next section

Human Capital Characteristics of Women The gender earnings differentials that persist in the United States and Ireland may result from differences in the human capital characteristics of women in the two

countries The following sections discuss the differences and similarities between the U.S and Ireland regarding women’s participation in the labor force, educational

attainment, and marital status

Female Labor Force Participation:

The female labor force participation rate in the United States is significantly higher than the corresponding level for Ireland In the U.S., nearly three out of four women participate in the labor force, but in Ireland, only one out of three participate There are two schools of thought on how the female labor force participation rate impacts the gender earnings differential First, if more women participate in the labor market, it follows that more barriers will have been removed and there will be more equality of wages Therefore, a higher female labor force participation rate should indicate a

narrower gender earnings differential However, the second school of thought suggests that with a lower female labor force participation rate, the women who work most likely have greater investments in education and thus would earn higher wages As more

women enter the labor market the educational qualifications begin to decrease and female average wages begin to decline which leads to a wider gender earnings differential In order to determine which scenario fits the U.S and Ireland, the differences in women’s educational attainment and marital status need to be developed

Educational Attainment of Women:

The majority of women who work full-time in both the U.S and Ireland

completed their formal education with secondary school In the U.S., a larger percentage

of women who work have completed college or advanced degrees than in Ireland; most

likely, this is because women in Ireland do not tend to invest in education to the same degree as women in the U.S because there is a greater expectation that women in Ireland will leave the work force once they are married and have children The U.S does not have the same expectation However, it is still a minority of women in both countries who have invested in higher levels of education; therefore, it is most likely that a

minority of women are earning relatively high wages Because the majority of working women have not obtained higher levels of education in both the U.S and Ireland, there is

an increased likelihood that the level of educational attainment of women would impede the reduction of the gender earnings differentials in the two countries However, it is important to note that the female rate of return to education is higher in the U.S than in Ireland, which should narrow the U.S gender earnings differential relative to Ireland The results indicate that both the U.S and Ireland experienced an increase in the rate of return to education for women between 1987 and 1993, which helped facilitate the

reduction in the gender earnings differentials that occurred in both countries

of the gender earnings differential (Callan, 1991) Married women tend to have children, which means that they have most likely interrupted their careers with time spent out of the labor force at some point which works to widen their gender earnings differential This issue of marital status and time spent out of the labor force is perhaps the most

significantly different factor between the U.S and Ireland The vast majority of women who work full-time in Ireland are single, whereas, in the U.S., most of the women who work full-time are married Therefore, the marital status factor should impede reductions

in the gender earnings differential in the U.S but should facilitate reductions in the

in the U.S are married and the majority of women who work in Ireland are single

The female labor force participation rate in the U.S is relatively high which

should be expected and is not hard to understand given the increasingly common

expectation that women will pursue professional careers in place of or in addition to family responsibilities In contrast, Ireland has maintained a relatively low female labor force participation rate mainly because Ireland is a small country with a strong Roman

Catholic tradition and strong traditional family values where the male head of household,

is gainfully employed outside the home to support his wife and family The Roman Catholic faith does not support divorce or the use of artificial contraceptives, and because Ireland is overwhelmingly Catholic, Irish families, in general, tend to stay intact and be relatively large As a result, there are few single parent families which require the

entrance of the female head of household, into the labor market in order to survive As such, the traditional situation for the vast majority of the Irish population is that the male

is employed outside the home and the female works at home duties

Summary The United States and Ireland both experienced relatively large gender earnings

differentials in 1987; however, the two countries are different with respect to the labor

market wage-setting institutions, the macroeconomic conditions, and the human capital characteristics of women The labor market wage-setting institutions in the U.S work to widen the gender earnings differential while in Ireland, the wage-setting institutions work

to narrow the differential The effect of an increasing growth rate coupled with a

decreasing unemployment rate is an increase in wages When average female wages increase more than average male wages the gender earnings differential narrows which is what happened in the U.S and Ireland between 1987 and 1993 The differences in the human capital characteristics of women in the U-S and Ireland are significant Women

in the U.S are obtaining higher levels of education than women in Ireland; however, it is still a minority of women that are pursuing college and advanced degrees in both

countries; therefore, it is unlikely that the level of educational attainment is working to

narrow the gender earnings differential in either country However, the rate of return to education for women is increasing in both countries between 1987 and 1993 which helps

to narrow the gender earnings gap

The differences in the marital status of women in the U.S and Ireland is also

significant More women who work full-time in the U.S are married, which tends to

widen the gender earnings differential, whereas, more women who work full-time in Ireland are single, which tends to narrow the gender earnings differential Finally, the differences in the amount of home time between women in the U.S and Ireland is

significant Women in the U.S do not tend to leave the labor force as a result of family responsibilities to the same extent as women in Ireland Regardless of the duration of time spent out of the labor force, whether it is temporary as in the U.S or sustained as in Ireland, the interruptions in a woman’s work experience tend to widen the gender

earnings differential The aim of this paper is to explore the many issues brought up in

this chapter in order to determine quantitatively what factors are impeding or facilitating the reduction in the magnitude of the gender earnings differentials in the U.S and Ireland.

The following review of the literature is organized according to each of the three models used in this paper First, the theoretical background that led to the empirical specification of Mincer’s human capital earnings function is presented followed by some highlights from the rate of return to education literature Second, the wage

decomposition literature provides the context for a discussion of the gender earnings differential followed by the literature on the Juhn, Murphy, and Pierce wage differential residual decomposition

The Rate of Return to Education The early development of human capital theory is based on the observation that a positive relationship exists between education and earnings which means that schooling is

an investment that is rewarded in the labor market (Willis, 1986, 535) The early work in human capital theory used statistical earnings functions of the form:

to express the relationship between earnings, y, and years of schooling, s However, the expression f (s, x) was simply a functional form that best fit the available data rather than

an incorporation of human capital theory into an empirical specification [such as in

Becker and Chiswick (1966)] Becker (1964) and Mincer (1958, 1962, 1974) were the pioneers in formulating such an empirical specification (Willis, 1986, 541) In fact,

Mincer’s human capital earnings function has been hailed as “‘one of the great success stories of modern labor economics” and has been used in “hundreds of studies using data

16

from virtually every historical period and country for which suitable data exist” (Willis,

1986, 526)

Human capital theory states that people invest in themselves and incur present costs for a future gain The rate of return to education provides an indication of how much of a return on one’s investment in education may be expected in terms of higher lifetime earnings and is found by relating the distribution of the logarithms of earnings to the human capital characteristics that effect earnings in the following way:

InY = a + By S+ By EX + B3 EX* + u (2)

Mincer’s human capital earnings function fits an ordinary least squares regression with the logarithm of earnings as the dependent variable and years of education, years of experience, and years of experience squared as the independent variables (1974) The coefficient on the years of schooling variable is interpreted as an estimate of the rate of return to one year of education The extended earings function is a step-function

specification that uses dummy variables to represent the different levels of education and, thus, enables the marginal effect of each subsequent level of schooling to be determined

(Psacharopoulos, 1994, 1326)

Some important conclusions have emerged out of the literature on the rate of return to education that are relevant to this paper First, the completion of secondary and tertiary levels of education result in significantly higher lifetime earnings (Blaug, 1987) Second, the rate of return to secondary school is, on average, higher than the return to a college or university education (Psacharopoulos, 1981) Third, the returns to education for women are higher than the returns for men (Psacharopoulos, 1994, 1327)

One aim of this paper is to compare the rates of return to educational attainment for the various levels of education within the American and Irish educational systems in order to highlight the affect of education on earnings Of particular interest is the

difference in the rate of return to education for men and women which relates to the

gender earnings differential, specifically, the differences in the rewards to the human capital characteristics of men and women in the U.S and in Ireland

The Gender Earnings Differential

The rate of return to education is an important link to the gender earnings

differential because while the return to education indicates how much of a return to a woman’s investment in education may be expected in terms of higher lifetime earnings, the gender earnings differential indicates how the wage structure of the labor market will reward her investments Education plays a critical role in the gender earnings differential

as evidenced by Blau and Beller’s finding that the decrease in the returns to education for women in the U.S in the 1970s widened the gender eamings differential more than any other factor considered (1988, 527)

Theoretically, the individual decision of how much education to invest in depends

on the rate of return to that amount of education This investment decision is important for women because it is linked to one’s participation in the labor market While the costs

of education are high, the return to education for women tends to be greater than the return to education for men (Psacharopoulos, 1994, 1327), which would suggest that women should invest in education and gain the returns in the labor market Additionally, labor force participation has been shown to be “strongly related to educational

attainment” as women with higher levels of education are more likely to participate in the labor force than women with lower levels of education (Killingsworth and Heckman,

1986, 123) Also, women with higher levels of educational attainment are less likely to interrupt their careers with time spent out of the labor force and, therefore, attain higher earnings, which leads to less of a differential than for women with lower levels of

educational attainment (Ferber and McMahon, 1979) While gender earnings

differentials are significant for both highly educated and less educated women (Ferber

and Kordick, 1978), there is a smaller gender earnings differential for women with more

education than for women with less education (Mincer, 1974) This point is illustrated

for the U.S as the female-male hourly earnings ratio in the 1960s and 1970s increased 11.4 percent for women with some college, but increased only 1.5 percent for women with less education (Fuchs, 1974)

The gender earnings differential occurs for a number of reasons One reason is that women are more likely than men to interrupt their careers with time spent out of the labor market over the course of their lifetimes due to childbearing and childrearing

responsibilities Another reason the gender earnings differential exists stems from the structure of wages and occurs because women tend to become occupationally segregated into typically female jobs Another reason the gender earnings differential exists is

because women do not tend to invest in educational attainment and job skills to the same

extent as men And finally, another reason the gender earnings differential exists is

because there is gender discrimination in the labor market The contribution of each to the gender earnings differential will be discussed in the following paragraphs

One significant factor that contributes to the explanation of the gender earnings differential is the difference in the length of time men and women spend in the labor

force Women’s wages are lower because, over the course of a lifetime, an average

woman spends less time in the labor force than an average man Employers are aware of this difference and may engage in “statistical discrimination,” whereby the employer is less likely to invest in on-the-job training if s/he suspects the employee will remain in the labor force for a relatively short time (Thurow, 1975) Asa result of childcare and family responsibilities, it is more likely that a female employee’s job tenure will be interrupted at some point in her career

Another factor that contributes to the explanation of the gender earnings

differential is the structure of wages, or the set of prices, that prevail in the labor market