ags data analysis the gender wage gap 2013 an analysis of the australian graduate labour market

AGS DATA ANALYSIS

THE GENDER WAGE GAP

2013AN ANALYSIS OF THE AUSTRALIAN GRADUATE LABOUR MARKET

EDWINA L INDSAY, GCA

2

MEDIA

3

• Prior to the ‘60s, males wages higher than female wages

due to familial obligations.

• National Wage Case, 1967

• Equal Pay Case, 1969

• 1984 Sex Discrimination Act, 2006 Work Choices, 2009 Fair

Work, 2012 Workplace Gender Equality legislation.

AUSTRALIAN POLITICAL FRAMEWORK

4

• Equal Pay Case, 1969

WOMEN DEMONSTRATING OUTSIDE MELBOURNE’S TRADES HALL IN SUPPORT

OF EQUAL PAY IN 1969.

5

• Gender wage gap increases as age increases

• Disparities in labour market experience

• Career breaks

• Hours worked

• Differences in level and field of education

• Occupational choices and Industry

• Region of employment

KEY CONTRIBUTORS

6

Graduate labour market

• Key contributors were ‘observed’ factors such as:

- Hours worked and field of education (females over-

represented in lower-earning fields of education) (Finnie

and Wannell, 2004)

- Industry of employment and field of education (males

more likely to be found in higher paying occupations)

(Jewell, 2008)

LITERATURE - INTERNATIONAL

7

• Broad labour market

- Borland, 1999 – 15 per cent

- ABS, 2014 – 17.1 per cent

• Graduate labour market

- Birch, Li and Miller, 2009:

- 2003 GDS data. Field of education, occupation type, and industry – a

gender wage gap of 3 per cent.

- Li and Miller, 2012:

- GDS data (1999 – 2009).

- Blinder- Oaxaca decomposition– a gender wage gap of 5 per cent.

LITERATURE - AUSTRALIAN

8

1. Investigates whether a gender wage gap exists within the

graduate population

2. The extent of the gender wage gap when the personal,

enrolment and employment characteristics of graduates are

held constant.

THE STUDY

9

• Graduate Destinations Survey (2013)

- 109,304 responses; a response rate of 60.0 per cent

- Reliability of GDS data (Guthrie and Johnson 1997)

• Sample restricted to:

- Australian bachelor degree graduates

- Aged less than 25

- In first full-time employment in Australia

- Indicated gender

- No missing data on key variables

DATA

10

• Dependent variable – annual starting salary

- Outliers excluded (below $20,000 and above $112,500)

• Final analysis sample of 8,185 graduates

- 3,103 males and 5,082 females

DATA

11

Figure 1: Distribution of full-time starting salaries for male and female graduates, 2013

DATA

12

OLS Regression

lnSi = β0 + βFi + βXi + εi

• lnSi = annual starting salary of graduate i expressed in

logarithmic form

• β0 = constant

• Fi = variable indicating that graduate i is female

• Xi = vector containing the various characteristics of graduate i

(including personal, enrolment and occupational characteristics)

• εi = an error term.

METHODOLOGY

13

Dummy variables

• Female

• Field of education (22)

• Personal and enrolment (4)

• State of employment (14)

• Other employment characteristics (6)

• Occupation (7)

METHODOLOGY

14

METHODOLOGYExplanatory Variables

Variable of interest Personal characteristics Employment characteristicsFemale Disability ¤ Weekly working hours

Omitted: Male Omitted: No disability Non-English speaking background Other employment characteristics

Field of education Omitted: English speaking background Small and medium enterprise Accounting Omitted: large enterprise

Agricultural Science Enrolment characteristics Public/government sectorArchitecture & Building Honours bachelor Omitted: private/not for profit sector

Art & Design Omitted: pass bachelor Short-term contract Biological Sciences Double degree Omitted: permanent or open-ended contractComputer Sciences Omitted: single degree Field of study of limited importance

Dentistry Omitted: field of study important/formal

requirementEarth Sciences State of employment In full-time work in final year of study

Economics & Business NSW CapitalOmitted: not in full-time work in final year of

studyEducation NSW Regional

Engineering VIC Capital OccupationLaw VIC Regional Managers

Mathematics QLD Capital ProfessionalsMedicine QLD Regional Technicians and Trades workers

Optometry SA Capital Clerical and administrative workersParamedical Studies WA Capital Sales workers

Pharmacy WA Regional Machinery operators and driversPhysical Sciences TAS Capital Labourers

Psychology TAS RegionalOmitted: Community and personal service

workers Social Sciences NT Capital

Social Work NT Regional Veterinary Science ACT

Omitted: Humanities Omitted: Regional South Australia

15

Model 1:

FINDINGS

• Controlling for no other factor, female graduates earn, on

average, 9.4 per cent less than male graduates.

• Aggregate 9.4 per cent gap is due to varying enrolment

patterns of males and females, and occupational pathways

resulting from these patterns.

Model 1

Female-0.094

(0.006)**

16

Model 2:

• Builds on Model 1 by controlling for field of education, personal

and enrolment characteristics.

• Female coefficient halved from -0.094 to -0.047.

• Field of education has considerable explanatory power on the

starting salaries of graduates.

FINDINGS

17

Model 2: Graduates average annual starting salaries: controlling for

gender and enrolment.

FINDINGS Model 1 Model 2 Model 2

Sex Field of education (cont.)

Female-0.094

(0.006)**-0.047

(0.006)** Medicine0.238

(0.021)**

Field of education (a) Optometry0.529

(0.060)**

Accounting 0.070

(0.014)** Paramedical Studies0.155

(0.012)**

Agricultural Science 0.069

(0.029)* Pharmacy -0.110

(0.020)**

Architecture & Building 0.061

(0.019)** Physical Sciences0.101

(0.034)**

Art & Design -0.121

(0.020)** Psychology0.026

(0.020)

Biological Sciences -0.002

(0.017) Social Sciences0.023

(0.029)

Computer Sciences 0.125

(0.019)** Social Work0.028

(0.032)

Dentistry 0.446

(0.052)** Veterinary Science 0.024

(0.048)

Earth Sciences 0.285

(0.033)** Personal characteristics

Economics & Business 0.059

(0.011)** Disability0.023

(0.016)

Education 0.177

(0.013)** Non-English speaking background

-0.003 (0.008)

Engineering 0.306

(0.013)** Enrolment characteristics

Law 0.152

(0.019)** Honours bachelor0.114

(0.010)**

Mathematics 0.134

(0.038)** Double degree0.107

(0.008)**Adjusted R2 .026 .203

Adjusted R2 .203

F-statistic 221.85 78.03 F-statistic 78.03Sample size 8,185 8,185

Sample size 8,185

18

What can explain the 9.4 per cent gap?

• Traditional gender patterns

• More males in higher paying fields.

• Engineering vs. Humanities

FINDINGS

19

Model 2 : Graduates average annual starting salaries: controlling

for gender and enrolment.



Female-0.094

(0.006)**-0.047

(0.006)** Medicine0.238

(0.021)**


(0.060)**

Accounting 0.070


(0.012)**


(0.029)* Pharmacy -0.110

(0.020)**



(0.034)**

Art & Design -0.121


(0.020)



(0.029)


(0.019)** Social Work0.028

(0.032)

Dentistry 0.446


(0.048)





(0.016)

Education 0.177


-0.003 (0.008)

Engineering 0.306


Law 0.152


(0.010)**

Mathematics 0.134


(0.008)**Adjusted R2 .026 .203

Adjusted R2 .203


Sample size 8,185

20

Table 1: Graduates’ field of education enrolment patterns, by gender, 2013 (%)

SAMPLE DESCRIPTIVES

Male Female Total Male Female Total

Gender 38.0 62.0 100.0 Field of education (continued)

Field of education Humanities 5.7 11.6 9.3

Accounting 9.4 6.6 7.7 Law 2.4 3.4 3.0

Agricultural Science 1.1 0.9 1.0 Mathematics 1.0 0.3 0.6Architecture & Building 4.0 2.1 2.8 Medicine 2.3 2.0 2.1

Art & Design 2.0 2.9 2.5 Optometry 0.2 0.2 0.2

Biological Sciences 3.1 4.4 3.9Paramedical Studies 6.3 21.0 15.4

Computer Sciences 6.0 0.8 2.8 Pharmacy 2.2 3.0 2.7

Dentistry 0.2 0.4 0.3 Physical Sciences 1.2 0.4 0.7

Earth Sciences 1.4 0.4 0.8 Pyschology 1.1 3.3 2.4Economics & Business 21.6 18.8 19.8 Social Sciences 0.6 1.3 1.1

Education 3.5 10.9 8.1 Social Work 0.2 1.3 0.8

Engineering 24.6 3.7 11.6Veterinary Science 0.0 0.6 0.4

21

Model 2:

• But – not all female-dominated fields are associated with lower

starting salaries.

• E.g. Education and Paramedical Studies.

FINDINGS

22



Female-0.094

(0.006)**-0.047

(0.006)** Medicine0.238

(0.021)**


(0.060)**

Accounting 0.070


(0.012)**


(0.029)* Pharmacy -0.110

(0.020)**



(0.034)**

Art & Design -0.121


(0.020)



(0.029)


(0.019)** Social Work0.028

(0.032)

Dentistry 0.446


(0.048)





(0.016)

Education 0.177


-0.003 (0.008)

Engineering 0.306


Law 0.152


(0.010)**

Mathematics 0.134


(0.008)**Adjusted R2 .026 .203

Adjusted R2 .203


Sample size 8,185

Model 2 : Graduates average annual starting salaries: controlling

for gender and enrolment.

23

Table 1: Graduates’ field of education enrolment patterns, by gender, 2013 (%)

FINDINGS

Male Female Total Male Female Total

Gender 38.0 62.0 100.0 Field of education (continued)

Field of education Humanities 5.7 11.6 9.3

Accounting 9.4 6.6 7.7 Law 2.4 3.4 3.0

Agricultural Science 1.1 0.9 1.0 Mathematics 1.0 0.3 0.6Architecture & Building 4.0 2.1 2.8 Medicine 2.3 2.0 2.1

Art & Design 2.0 2.9 2.5 Optometry 0.2 0.2 0.2

Biological Sciences 3.1 4.4 3.9Paramedical Studies 6.3 21.0 15.4

Computer Sciences 6.0 0.8 2.8 Pharmacy 2.2 3.0 2.7

Dentistry 0.2 0.4 0.3 Physical Sciences 1.2 0.4 0.7

Earth Sciences 1.4 0.4 0.8 Pyschology 1.1 3.3 2.4Economics & Business 21.6 18.8 19.8 Social Sciences 0.6 1.3 1.1

Education 3.5 10.9 8.1 Social Work 0.2 1.3 0.8

Engineering 24.6 3.7 11.6Veterinary Science 0.0 0.6 0.4

24

Model 3:

• Builds on Models 1 and 2, by adding occupation and

employment characteristics.

• The addition of the various employment variables in Model 3 only

changed the female coefficient marginally, from -0.047 to -0.044.

• Adjusted R2 of .344

• 4.4 per cent figure is similar to previous findings: 3 per cent by

Birch, Li and Miller (2009) and 5 per cent by Li and Miller (2012).

FINDINGS

Model 1 Model 2 Model 3

Female-0.094

(0.006)**-0.047

(0.006)**-0.044

(0.006)**

25

1. Field of education characteristics of graduates assert considerable explanatory

power

- Differences in male and female enrolment patterns

- Field of education controls halved female coefficient

2. After controlling for all explanatory variables, gender wage gap of 4.4 per cent

remained unexplained by our data.

- Differences not captured in our data/models.

- Differences in negotiating behaviour?

- Discriminative practices within the workplace?

- Need for social reform?

- Female participation in STEM subjects?

- Need for further research – perhaps using a matching technique and analysing longitudinal

data (BGS).

CONCLUSIONS

26

MEDIA

27

QUESTIONS?

An analysis of the gender wage gap in the Australian graduate labour market, 2013

[email protected]

Thank you.

mailto:[email protected]

ags data analysis the gender wage gap 2013 an analysis of the australian graduate labour market

Documents