- Progress and Preliminary Outcomes Prof. Dr. Seo-Young Cho Faculty - - PowerPoint PPT Presentation

InGrid2 Research Visit

Project on Female Labor and Gender Discrimination in STEM fields

• Progress and Preliminary Outcomes
• Prof. Dr. Seo-Young Cho

Faculty of Economics Philipps-University of Marburg, Germany HIVA KU-Leuven, 30th March, 2018

Main Focus and Research Questions

• This project addresses female underrepresentation in STEM fields

that is still significant in most countries despite noticeable improvement in female education and employment over the last decades.

• This project aims to identify systematic gender inequality in STEM

fields:

• Whether gender pay gaps in STEM fields are larger than in non-

STEM fields.

• Whether gender gaps in social capital – trust, relationship, network,
• etc. – are greater in STEM fields than non-STEM fields.

2

During My Research Visit

• Data examination and organization
• European Working Conditions Survey (EWCS)
• Discussions on the concepts and data with HIVA researchers
• Participation in Workshop Education Economics at Economics

Faculty, KU Leuven

• Preliminary regression analysis

3

Hypotheses for Empirical Investigation

• Gender effect
• H1. The gender effect on wage/perceived discrimination/ trust in workplace/

network participation is negative for female workers.

• STEM-specific gender effect
• H2. The negative gender effect is greater for female workers in STEM fields

compared to others in non-STEM industries.

• Gender-matching effect
• H3. Gender-matching environments reduce the negative gender effect.

4

Model

Outcomei = β1genderi + β2STEMi + β3genderi*STEMi + β4gender/bossi + β5gender-matchingi + Xiψ + Ziω + + countryiƝ + ui

• Sample: full-time employed (unemployed, part-time, self-employed are

excluded in the sample)  to minimize self-selection into employment types

• i = {1,,,,, 34,370 }, individual workers
• Outcome vector

= {wage; trust in colleagues/management; perceived discrimination – gender, age, and ethnic; fairness in workload and evaluation; training opportunities; cooperation with colleagues/bosses; experience of being abused at work}

5

Model (cont.)

• Gender: the gender of employee i (1 = female, 0 = otherwise)
• Gender/boss: the gender of the direct boss of employee i
• Gender-matching: gender of employee i * gender of the boss
• STEM: industrial classification indicating STEM intensity
• Vector X: individual characteristics (education level, household size,

marital status, age, years of tenure, tasks, etc.) – controlling factors

• n the supply side
• Vector Z: firm characteristics (firm size, location, main production,

years, etc.) – controlling factors on the demand side

• Country: country fixed effect (35 countries)

6

STEM Identification (Measurements)

• Approach 1 (US Brooking Institute classification)
• Constructing STEM dummies

extended: architecture and engineering + computer and math + life and health limited: : architecture and engineering + computer and math

• Approach 2 (UK Commission for Employment and Skills

classification)

• STEM intensity based on proportion of people with STEM

qualifications (SOC composite score)

7

How much are you treated fairly at workplace?

• probit

ME

• Rob. Std. Err.

z P>z gender

• .0844422

.0159609

• 5.29

0.000 boss_gender .0514125 .0249771 2.06 0.040 gender#boss_gender .0550371 .0301255 1.83 0.068 STEM .1520256 .0258448 5.88 0.00 gender#STEM Controls (X) Controls (Z) Country FE

• .1627088

.03581 Yes Yes Yes

• 4.54

0.000

8

How much does your boss give you praise and recognition when you do a good job?

9

• probit

ME

• Rob. Std. Err.

z P>z gender

• .0634189

.0154391

• 4.11

0.000 boss_gender .1293307 .0245852 5.26 0.000 gender#boss_gender .033078 .0296301 1.12 0.264 STEM .1671826 .0252718 6,62 0.000 gender#STEM Controls (X) Controls (Z) Country FE

• .1850912

.0360024 Yes Yes Yes

• 5.14

0.000

Have you taken training (paid by employers) over the past 12 months?

Probit ME

• Rob. Std. Err.

Z P>z Gender

• .0081336

.0183569

• 0.44

0.658 boss_gender .234772 .0282059 8.32 0.000 gender #boss_gender .0847273 .0341452 2.48 0.013 STEM .4770867 .030206 15.79 0.000 gender#STEM Controls (X) Controls (Z) Country FE .0230659 .0423598 Yes Yes Yes 0.54 0.586

10

Log Wage

OLS Coef.

• Std. Err.

t P>t gender

• .2698182

.0214033

• 12.61

0.000 boss_gender

• .0397034

.0345846

• 1.15

0.251 Gender #boss_gender

• .0238794

.0417666

• 0.57

0.568 STEM .4614581 .0371782 12.41 0.000 gender#STEM Controls (X) Controls (Z) Country FE .1145812 .0526771 Yes Yes Yes 2.188 0.030

11

Some preliminary findings

• STEM provides positive working environments for men
• Male workers perceive higher levels of fairness and recognition, earn more

and receive more training opportunities in STEM fields, compared to men in non-STEM fields.

• The effect of STEM is mixed for female workers
• Negative effect on non-monetary conditions: female workers perceive lower

levels of fairness and recognition in STEM fields, compared to other female workers in non-STEM fields.

• Positive effect on earnings and job training & the positive effect of STEM on

earning is greater for females than males!

12

Some preliminary findings (cont.)

• Gender inequality in STEM fields is more related to non-monetary

forms of perceived discrimination and working relationship than monetary ones.

• Positive gender-matching effect
• Female workers perceive higher levels of fairness and receive more

training opportunities if they work with female bosses. Gender inequality in STEM fields can be reduced by promoting women into lead-positions.

13

Next Steps

• Extension of data
• Different waves of EWCS
• EU Labor force Data (incorporating detailed education/degrees

variables and employees’ intention to leave)

• Causality issues: self-selection into STEM fields due to unobserved

gender-based characteristics

• PSM
• Structural Simultaneous Equations Model (wage-training-trust-

networks)

14