Is Casual Employment a Bridge or a Trap

1
Is Casual Employment a Bridge or a Trap?

• Jenny Chalmers,
• National Drug and Alcohol Research Centre
• University of New South Wales
• C. Jeffrey Waddoups,
• Department of Economics
• University of Nevada Las Vegas

2
Definition of Casual Employment

• Casual Work form of employment in which workers
  do not have access to annual leave, paid sick
  leave, paid public holidays, notice of dismissal,
  and redundancy pay
• ABS Definition (used in HILDA and our study)
  employees without access to either paid holiday
  or paid sick leave.

3
Background

• The incidence of casual employment has increased
  from 16.0 percent in 1985 to 26.9 percent in 2006
• Benefits and costs associated with increased
  incidence of casual employment .
• Reduced opportunity for meaningful on-the-job
  human capital development
• Reduced opportunity for career advancement
• Flexibility in the labour market, a potential
  benefit in a dynamic market economy.
• Flexibility for workers to meet household
  obligations

4
Background

• The debate is not unique to Australia. As
  nations labour markets liberalise, questions
  about the trade-offs between flexibility and job
  quality naturally arise.
• For example, Farber (1999) in the U.S., Natti
  (1993) in the Nordic countries, Booth, Fancesconi
  and Frank (2002) on the UK, Chalmers and Kalb
  (2001) and Gaston and Timcke (1999), Productivity
  Commission (2006) in Australia.

5
Research Questions

• We use HILDA to address nature and determinants
  of mobility from casual employment
• Does time spent in casual employment increase
  (bridge) or decrease (trap) the likelihood of
  transition to permanent employment?
• Although casual employment is too complex to be
  characterised as wholly a bridge or a trap,
  we can still assess aspects of mobility that are
  consistent with the two metaphors.

6
If Casual Employment is a Bridge, then one
would expect to see

• Job training
• Development of work discipline
• Development of networks to enhance mobility to
  permanent job
• The above should result in increasing probability
  of exit to permanent employment as tenure in
  casual employment increases.

7
Public Policy If Bridge

• If casual employment takes on the trappings of a
  bridge, having relatively large numbers of jobs
  without standard entitlements or protections is
  not particularly costly to society, suggesting
  policy intervention less likely to be warranted.

8
If Casual Employment is a Trap, then one would
expect to see

• Involuntary employment in sub-standard jobs
• Perhaps development of negative human
  capital(scarring)
• Little or no relevant training
• Negative signals about potential productivity to
  employers offering permanent jobs
• The above should result in reduced probability of
  transition to permanent jobs with increased
  tenure in casual employment

9
Public Policy if Trap

• If casual employment acts like a trap, having
  relatively large numbers of casual jobs would be
  relatively costly to society, which indicates
  that policy intervention to limit casual
  employment is more attractive.

10
Description of the Data

• HILDA
• Four Waves 2001 2004
• Employees (not self-employed) in all four years
• Ages 15- 64 in 2001.
• Contain observations on all relevant variables
• Two Samples
• Sample One observe those who report casual
  employment in 2001.
• Sample Two observe those that commence casual
  spell in 2002.

11
Sample One Casual in 2001

• Sample one
• Casual in 2001 (left truncation, we do not know
  how long workers were in casual employment before
  2001).
• Count the number of periods in a casual job
  starting in 2001.
• Example if casual in 2001, casual in 2002, then
  permanent in 2003, it implies two periods (years)
  of casual
• Reporting once per year implies unobserved
  transitions could have occurred

12
(No Transcript)
13
(No Transcript)
14
Sample Two Started Casual Spell in 2002

• Sample Two
• Started casual spell in 2002.
• Originate from non-employment or permanent
• No left truncation, but fewer years to observe
  mobility

15
(No Transcript)
16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
Accelerated Failure Time (AFT) Model
ln (ti) Xiß ei

• tduration in casual spell
• Xvector of covariates
• ßa vector of parameters
• ea vector of error terms
• Error terms assumed to be distributed as a
  Weibull and log-logistic

23
Weibull and Log-Logistic

• Weibull restricts hazard (conditional
  probability of exit from casual) to be
  monotonically increasing or decreasing
• Log-Logistic Allows for increasing then
  decreasing hazard (inverted U-shaped)

24
Shape of Hazard Bridge or Trap

• An increasing hazard suggests that probability of
  exit at time t, given survival until time t, is
  increasing (Bridge)
• A decreasing hazard suggests that probability of
  exit at time t, given survival until time t, is
  decreasing (Trap)

25
Covariates in an AFT Model

• Model controls for variables generally thought to
  affect labour market outcomes (human capital,
  etc).
• Covariates essentially shift the hazard, and
  either
• accelerate the time to failure if a negative
  value (a shorter duration in casual)
• or
• Decelerate the time to failure if a positive
  value (a longer duration in casual)

26
Estimated Models

• STATAs STREG used to estimate Models
• Predicted Durations
• Sample 1 Med3.4 years, Mean5.1years
• Sample 2 Med2.9 years, Mean4.0 years

27
Parameter Estimates

• Statistically Significant Estimates,
• Tenure with current employer ()
• Part-Time ()
• Female (, Weibull sample 2)
• Children ages 0-4 (-, Sample 2)
• Married (-, sample 1)
• Large Employer (-, Weibull sample 1)
• To casual from permanent (-, only relevant to
  Sample 2)
• Model Specific parameters (Weibull p,
  Log-logistic gamma)

28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
Other Results

• Interacted a dummy variable controlling for
  female with tenure, part-time, married, and
  children dummies.
• Only statistically sig. result (Sample 1,
  children ages 0-4, sign was negative).

33
Which is Preferred, Weibull or Log-Logistic?

• Akaike Information Criterion
• AIC -2(Log-likelihood) 2(kc),
• knumber of covariates
• cnumber of model-specific paramters
• the model with the lower value of the AIC is
  preferred model.

34
(No Transcript)
35
Conclusions

• Duration in casual jobs explained by
  circumstances of present employment (tenure with
  current employer, part-time)
• Do not observe whether this is a sign of relative
  comfort with casual employment, or involuntary
  lack of mobility
• Duration also explained by household
  circumstances
• Marriage and young children linked with a
  decrease in duration
• No apparent systematic use of casual employment
  for flexibility in attending to household duties.
  In fact household duties correlated with
  permanent employment.

36
Conclusions

• Roughly 40 percent of casuals in 2001 who
  reported employment all four years did not make a
  transition to permanent work.
• Over 50 percent of casuals who commenced a casual
  spell in 2002 still reported casual employment in
  2004.
• Roughly 60 percent (sample 1)and 50 percent
  (sample 2) do make the transition.
• It is still difficult to determine whether a
  casual spell facilitated, or delayed, such a
  transition.

37
Conclusions

• Inverted U-shaped hazard suggests that the length
  of time in casual decreases probability of exit.
  
• Question about whether this is because of
  trapping or unobserved heterogeneity.
• On its face, casual employment would seem too
  complex to characterize as wholly a bridge or a
  trap, and our mixed results confirm such
  intuition.