Database Reference
In-Depth Information
The working context of
professional sports
, such as baseball, basketball, foot-
ball, and soccer, offers an unusually good opportunity of studying discrimination.
The problem of estimating the productivity of workers is here substantially solved
by extensive, publicly available (from online sport almanacs), measures of the per-
formances of players and coaches. Research has covered discrimination in hiring, in
retaining (along seasons), in segregating (to specific game roles), and in salary of
players, as well as customer discrimination. The last topic is also known as
fan
discrimination
, typically measured using TV audience (Aldrich et al., 2005), game
attendance (Foley & Smith, 2007; Hersch, 2009; Wilson & Ying, 2003), the trad-
ing value of sport cards (Broyles & Keen, 2010; Primm et al., 2011), the votes for
best player awards (Jewell et al., 2002). As far as salary discrimination in profes-
sional sports is concerned with, there is an extensive literature on the subject. We
mention only a few recent papers (Berri & Simmons, 2009; Holmes, 2011; Frick
& Deutscher, 2009; Goddard & Wilson, 2009; Palmer & King, 2006; Yang & Lin,
2010), and refer the reader to the surveys (Kahn, 1991b, 2000, 2009).
Extensions of taste-based discrimination, called
search models
(Altonji & Blank,
1999; Lang & Lehmann, 2011), take into account the costs for workers of searching
jobs by interacting with prejudiced and non-prejudiced firms, and, for consumers,
the costs of searching sellers of their same racial group (Flabbi, 2010; Kuhn & Shen,
2009; Sulis, 2007; Usui, 2009). Finally, a line of studies, initiated by (Hamermesh &
Biddle, 1994), investigates the “beauty premium” in labor market. As a recent work,
we mention (Cipriani & Zago, 2011), who study favoritism to attractive students in
taking exams at University. The effectiveness of blind decisions in reducing gender
discrimination has been evaluated for orchestra auctions in (Goldin & Rouse, 2000).
Approaches on statistical discrimination.
Some extensions of the statistical dis-
crimination model deal with what happens as the employer's information on work-
ers' productivity changes, e.g., at the selection time or over the course of the job.
These dynamic extensions, contrasted to a static model, are known as
employer
learning
models. (Farber & Gibbons, 1996) propose a dynamic model of learning
about worker ability in a competitive labor market. Altonji and Pierret provide a
first important strand literature on learning models (Altonji & Pierret, 2001). We
complement the studies surveyed in the recent paper (Lang & Lehmann, 2011) by
mentioning: (Cheung, 2010), in testing whether parental education is used as a proxy
for the ability of workers; and (Wang, 2010), in considering height as an easily ob-
servable characteristic.
Also, the differential observability or learnability of worker's productivity among
groups has been taken into account by
screening discrimination
models, originally
introduced in (Lang, 1986). Such differences are due, e.g., to miscommunication
problems or weak interactions among groups. As an example, (Grogger, 2011) an-
alyzes audio data from telephone interviews to understand the role that speech may
play in explaining racial wage differences, and (Pinkston, 2006) shows that the level
of education has a large impact on wages. Similar work emerges from the health
literature, when testing whether miscommunication problems influence a diagnosis
Search WWH ::
Custom Search