Inequality across race and gender

Question 1

What is the definition of discrimination?

Answer 1

Discrimination is when members of a ‘minority’ group are treated differently than members of a ‘majority’ group with identical productive characteristics

Question 2

While there is evidence of an indication of discrimination why is it not taken as conclusive? (3)

Answer 2

Unobserved factors:

• Productivity can depends directly on f (e.g. beauty)
• B (returns to x) could be endogenous (do we reward male productivity characteristics more than female)
• x’s themselves could be endogenous (expectation of future discrimination reduces x’s)

Question 3

Which audit study looks at discrimination?

Answer 3

Bertrand and Mullainthan (2004)

-Are Emily and Greg more employable than Lakisha and Jamal?

Question 4

which natural experiment looks at discrimination?

Answer 4

Goldin and Rouse (2000)

-orchestrating Impartiality

Question 5

What are the results from Bertrand and Mullainthan Study? (4)

Answer 5

1. White name received 50% more call-backs
   - Equivalent to 8 more years experience
2. Resume quality has a greater effect for white names
   - If AA have higher skills the call-backs increase but not as much as it does for whites
3. No evidence that AA benefit from living in whiter, more educated, post codes.
4. Race gap relatively uniform across occupations, job requirements, employer characteristics and industries.

Question 6

Advantages of Bertrand and Mullainthan Study? (2)

Answer 6

They address the two main limitation of the audit study:

1. Use resumes instead of actual auditors so the participants cannot influence the results
2. Plus the cost of sending out resumes is far lower so they can collect more observations

Question 7

Limitations of Bertrand and Mullainthan Study? (7)

Answer 7

1. We are interested in actual job offers and wages, not just call backs.
2. Race is not directly reported and is inferred from the name.
3. Use very distinct names which may not represent the average applicant.
4. data is only from 2 of the large metropolitan cities.
5. limited occupation categories
6. No information on who else applied
7. No information on the employers

Question 8

What is the research design of ‘Are emily and greg more employable then Lakisha and Jamal?

Answer 8

Send fictitious resumes with randomly designed AA or white sounding names to job ads in Chicago and bostan

Question 9

How many employment adds did the Bertrand and Mullainthan Study reply too?

Answer 9

1300

Question 10

What is the research design of the Goldin and Rouse (2000) study?

Answer 10

The study attempts to isolate the importance of gender preference in a market setting

Question 11

What method does the Goldin and Rouse (2000) study take?

Answer 11

difference-in-differences

Question 12

What data did the Goldin and Rouse (2000) study use?

Answer 12

• Collected data from 8 major symphony orchestras
• Dating from 1950s to 1995
• Sample of auditions consists of 14, 121 person-rounds, 7,065 individuals and 588 audition-rounds

Question 13

What were the findings of the Goldin and Rouse (2000) study? (2)

Answer 13

1. Blind auditions during the preliminary rounds increased the likelihood that a female musician would advance to the next round by 11 percentage points
2. In the final round, blind auditions increased the likelihood of female musicians being selected by 30%

Question 14

Cautions of the Goldin and Rouse (2000) findings? (2)

Answer 14

1. Women who choose to audition for orchestras using blind auditions could be substantially different from those who choose to take part in non-blind.
2. Performers might perform different with/without blind screens.

Question 15

What are the four models for discrimination? (4)

Answer 15

1. taste for discrimination
2. Statistical discrimination
3. Overcrowding model
4. Institutional model

Question 16

Describe taste for discrimination?

Answer 16

When people act as if there is a non-pecuniary cost of associating with a particular group.

The source can be employer, employee or customer

Question 17

Describe statistical dsicrimination

Answer 17

Employment decisions are based not on individual productive characteristics but (perceived or accurate) group differences

Question 18

Describe the overcrowding model of discrimination

Answer 18

In the overcrowding model, minority groups crowd int o lower wage jobs either due to preference or lack of alternative opportunity

Question 19

Describe the institutional model for discrimination

Answer 19

Everyday operation of firms (which are controlled by majority groups) cause differences in productivity, promotion opportunities and pay.

Question 20

Which theoretical model for discrimination underpins the Bertrand and Mullainthan Study?

Answer 20

Employers located in areas which a higher % of black residents discriminate less, so there is some evidence of employer taste based discrimination

Question 21

Which theoretical model of discrimination underpins the Orchestrating Impartiality study?

Answer 21

We can NOT tell whether the evidence present is for taste based or statistical discrimination