Database Reference
In-Depth Information
and present discrimination by providing opportunities to those traditionally denied
for (ENAR, 2008; Holzer & Neumark, 2004; Sowell, 2005). They range from the
mere encouragement of under-represented groups to preferential treatment or quo-
tas in favor of those groups (see e.g., Holzer & Neumark, 2006; R. Lerner & Nagai,
2000).
Since discrimination can arise only through the application of different rules or
practices to comparable situations or of the same rule or practice to different sit-
uations, a relevant legal distinction is between direct and indirect discrimination.
When such rules or practices explicitly treat one person less favorably on a forbid-
den ground than another is, has been or would be treated in a comparable situa-
tion, we have
direct discrimination
, sometimes called
systematic
discrimination or
disparate treatment
. When an apparently neutral provision, criterion or practice re-
sults in an unfair treatment of a protected group, we have
indirect discrimination
,
sometimes called
adverse impact
(Tobler, 2008). While direct discrimination is in-
tentional and “directed” towards individuals, typically on the basis of their visible
traits, such as ethnic origin, race, sex and age, indirect discrimination is concerned
with avoiding the circumvention of the prohibition to discriminate, and to enforce
such a prohibition substantively, even in the case of unintentionality.
6.3
Labour Economic Perspective
In the labor market, different treatments among groups of workers can be measured
in terms of their wages (
wage differentials
), in the degree of participation in the
labor force (
employment differentials
), or in the degree of segregation in specific oc-
cupations or industries (
segregation differentials
). Public surveys routinely collect
data on demographic characteristics and attitudes of residents (e.g., in the US, the
General Social Survey - GSS), on the distribution of labor forces in the labor market
(e.g., Current Population Survey - GPS), and so on. Empirical research techniques
have applied statistical inference to collected data either with the purpose of test-
ing the consequences predicted by a theoretical economic model, or to assess the
contribution of different types of discrimination to the overall different treatments
in the labor market. The main data analysis techniques adopted include statistical
tests on rates and proportions (Agresti, 2002; Fleiss et al., 2003; Sheskin, 2004),
(generalized) linear regression models (Dobson & Barnett, 2008; Hardin & Hilbe,
2007; McCullagh & Nelder, 1989), and econometric models (Greene, 2008).
Two major theoretical models of discrimination have been considered in the eco-
nomic literature.
Taste-based discrimination
, originally proposed by (Becker, 1971),
has no rational or economic basis, but only a prejudiced personal taste against pro-
tected groups. Wage differentials are due to an additional psychological cost for
employing minority workers. Differently,
statistical discrimination
, originated by
(Arrow, 1971) and (Phelps, 1972) and systematized by (Aigner & Cain, 1977), starts
from the assumption that employers cannot perfectly assess worker productivity at
the time of hiring. This market imperfection gives them an incentive to use easily
Search WWH ::
Custom Search