What are the causes of age discrimination in employment?

1
What are the causes of age discrimination in
employment?

• Mathias Hungerbühler
• FUNDP Namur

2
Outline

• Definition and measurement of discrimination in
  economic theory
• Economic theories of discrimination
• What explains diffrences with respect to age in
  labour market outcomes?
• Is there age discrimination in labour markets?

3
Definition

• There is discrimination if irrelevant factors
  affect the (economic) outcomes.
• e.g. does race/gender affect wages of
  individuals, hiring or firing decisions of
  employers?
• What are relevant factors and which factors are
  considered irrelevant?

4
Measurement of discrimination
relevant factors that might explain
differences, e.g. education, experience,
outcome to be explained, e.g. wage or employment
rate
irrelevant factor, e.g. race, gender,
error term
coefficients to be estimated
5
Caveats of the measurement

• What are the explanatory variables? (Is age an
  explanatory variable?)
• Discrimination in explanatory variables (e.g.
  education)
• Endogeneity problems (e.g. is education lower
  because there is discrimination?)
• discrimination or unexplained differences?
  (problem of ommitted variables)

6
The taste theory of discrimination

• Employers might have a taste for discrimination
  of certain groups of the population
• e.g. some employer might pay black people
  constantly less than white people
• by discriminating people, the employer reduces
  its profits
• hence, if markets are competitive, this kind of
  discrimination is not sustainable! (but what if
  markets are not competitive?)

7
The taste theory of discrimination

• consumers might have a taste for discrimination
• e.g. you want to be served in a bar by a young
  women rather than an old man
• hence, a young woman is more productive in this
  job, simply because customers discriminate
• this kind of discrimination is thus sustainable
  in competitive markets

8
Statistical discrimination

• The employer cannot observe the caracteristics of
  an individual, but only the caracteristics of the
  group to which this individual belongs
• e.g. an employer cannot observe whether a woman
  is going to spend time off job for childcare, but
  he knows that on average, women are more often
  off the job to care their sick children than men
• Hence, the employer bases his decisions on the
  groups characteristics and discriminates
  individuals

9
Statistical discrimination

• Even though statistical discrimination is not
  efficient, it is sustainable in an economic
  equilibrium
• The main problem is that information is private
  to individuals and there is no possibility for
  credible commitement
• e.g. a woman cannot credibly commit to not having
  children

10
Hiring discrimination in imperfect labor markets

• For a vacancy, an employer might get several
  applications of equally qualified individuals
• He might then discriminate because he has some
  taste for it
• This kind of discrimination is sustainable
• It predicts that the discriminated group has both
  lower wages and lower employment (higher
  unemployment)

11
Differences in labour market outcomes with
respect to age

• wages
• hiring
• promotion
• training / continuous education
• firing

12
 Standard  factors that are able to explain
part of these differences

• education
• experience
• sector of employment
• etc.

13
Forward-looking employers and differences

• An old individual will most probably stay less
  time in the firm than a young individual
• A rational employer takes account of this when it
  comes to hiring, firing, etc.
• e.g. he rationally chooses to hire a younger
  individual
• Is this discrimination? Most economists would
  answer negatively i.e. they consider that age is
  a relevant factor in this case

14
and its policy response

• to soften the situation of the old, the
  government has set up generous pre-retirement
  plans that are partly publicly financed
• Though this softens the situation of the old, it
  also gives incentives to the firm to fire old
  people rather than the young. There are thus
  perverse effects of this policy measure.

15
Wages and hiring

• Are wages of the old  too high ?
• Incentives to set wages of the old high given by
  pension plans
• e.g. pensions are calculated on the last five
  years of contribution. Hence, employers and
  unions agree to set the olds wages at a  too
  high  level
• Thus, an employer will hesitate to hire an old
  individual and their probability of being fired
  increases.

16
Attachement of old people

• Old individuals might less mobile across
  geographical regions and economic sectors
• e.g. an old might own his house, which will
  decrese his geographical mobility
• hence, he will have more difficulties to find a
  new job, and he can less threaten his current
  employer that he might leave his job if he gets a
  better offer

17
Individual choices?

• Unemployed old individuals have more difficulties
  to find a new job than young people
• but if they find a new job, their wage is
  higher than the wage for the young individuals!
• Are old people simply more selective in job
  applications and acceptance?

18
Is there age discrimination in labour markets?

• Are the previous-mentionned mechanisms
   discriminatory ?
• It is almost impossible to measure age
  discrimination in the usual way discrimination is
  measured in economics, since age is certainly
  also an explanatory variable
• Is there taste discrimination? Statistical
  discrimination? Hiring discrimination?

19
Conclusions

• There are differences with respect to age in
  labour market outcomes
• There are plenty of non-discriminatory factors
  that are able to explain part of these
  differences
• Standard econometric techniques are not able to
  state whether there is age discrimination or not.
  Case studies might be more informative.
• Hence, economists can unfortunately not say much
  about the existence or not of age discrimination