Unemployment Compensation and the Risk of Unemployment The Case of Argentina

1
Unemployment Compensation and the Risk of
UnemploymentThe Case of Argentina
Ana Lucía Iturriza (ARGENTINA)
JJ/WBGSP Institute of Social Studies, 2005
2
The Topic

• As a JJ/WBGSP student of Economics of Development
  
• I was interested in choosing a topic related to
• MDGs
• Poverty Reduction
• The functioning of labor markets is a crucial
  determinant for poverty reduction.
• Income generation / distribution

3
The Context

• Size of the labor market (2003)

• Poverty and Unemployment (2000-03)

4
Compensation Programs

• Unemployment Insurance, in place since 1991.
• Unemployment Assistance (Plan Jefes), in place
  since 2002.

5
Research Question

• I test the standard prediction in labor
  economics.
• Does unemployment compensation affect the
  incidence or the duration of unemployment?
• If so gt the programs may be creating
  disincentives to reenter the employment.
• Particularly relevant for an assistance program
  such as Plan Jefes.
• I discuss whether to consider Plan Jefes
  participants into the employment (broad
  definition of employment) or not (restrictive
  definition of employment).

6
The Data

• Permanent Household Survey (EPH)
• Demographic indicators (age, sex, marital status,
  position in the household, region of residence)
• Labor market variables (condition of activity,
  time related variables, characteristics of last
  occupation when relevant)
• Benefit related variables (unemployed receiving
  insurance, assistance or condition of applicant).
• Panel merging two rounds October 2002 and May
  2003.
• Transition matrix.
• 21,000 observations of people in working age.

7
Empirical Strategy

• Econometrics to estimate the impact of Insurance
  and Assistance on the probability of transiting
  from UNEMPLOYMENT to EMPLOYMENT
• LOGIT MODELS gt Incidence - Probability
• COX MODELS gt Duration - Hazard rate
• NON PARAMETRIC MODELS gt Empirical pattern of
  exits from unemployment.
• Analysis performed for two different scenarios.
• Plan Jefes participants are employed (broad
  definition of employment)
• Plan Jefes participants are unemployed
  (restrictive definition of employment).

8
Example of the Data Used
9
Example of the Results Obtained
10
Interpretation of the Results

• Benefit related variables
• Unemployment insurance Negative but not
  significant.
• Unemployment Assistance Negative and
  significant. Compared to non-assisted workers,
  the probability of transitioning to employment is
  31 percentage points lower and the hazard rate
  is more than twice as less among the assisted.
• Condition of Applicant Negative but not
  significant.
• Other regressors that predict the transition
• Age, sex, position in the household, region of
  residence.

11
Conclusions

• Unemployment insurance
• It was not found to affect the employment
  behavior of participants.
• Assistance program (Plan Jefes)
• Under the broad definition of employment Plan
  Jefes favors reinsertion into the employment.
• Under the restrictive definition of employment,
  strong demotivating effects associated with the
  use of Plan Jefes are found.
• Policy instruments should be reviewed in
  accordance with the objectives for which they are
  created, particularly distinguishing between
  employment creation and other forms of social
  protection.

12
Drawbacks

• Plan Jefes creates obvious moral hazard problems
  (through its stipulation that the duration of
  unemployment assistance is unlimited).
• Conclusion insufficiently developed.
• Lack of clarity regarding the policy
  implications and recommendations derived from the
  study.
• The effects of the crisis on the labor market
  should be more carefully considered.
• The analysis could be improved with recent
  contributions on Argentinean economic crisis and
  recovery, for instance Calvos paper.