RDP 2020-05: How Risky is Australian Household Debt? Appendix B: Fit of Predictive Unemployment Model

Jonathan Kearns, Mike Major and David Norman

August 2020

Table B1: Marginal Effects of Probit Model
Dependent variable: probability of becoming unemployed
  SIH SIH HES
Previously unemployed 44.5***   46.2***
Earns rental income −4.3** −7.8*** −5.2**
Has housing debt −4.7*** −8.9*** −4.1***
Married −6.1*** −11.9*** −6.7***
Long-term health condition 6.7*** 12.0*** 6.5***
Age −0.3*** −0.4*** −0.3***
Born in English-speaking country −6.0*** −9.3*** −5.3***
Lives in capital city 1.1 0.8 2.2**
Educational attainment
Completed Year 12 −3.9*** 0.0*** −2.9**
Diploma −2.2*** −4.9*** −1.9***
University −5.7*** −10.8*** −4.5***
Family structure
Couple with dependent children 4.0** 4.3** 5.7**
Couple without dependent children 2.5 3.5 4.9***
Single with dependent children 2.4*** 4.1** 5.0***
Number of observations 20,785 20,785 11,367
Pseudo R2 0.18 0.11 0.19
Area under ROC curve 0.80 0.76 0.80

Note: *, ** and *** denote statistical significance at the 10, 5 and 1 per cent levels, respectively