Skip to main content

Table 10 State dependence in overeducation, women

From: Overeducation among immigrants in Sweden: incidence, wage effects and state dependence

  Random effects probit, 2002 − 2008
  Natives Western immigrants Non-Western immigrants
  III I II III I II III
Overeducation t-1 2.725*** 3.057*** 3.057*** 2.891*** 3.298*** 3.299*** 3.101***
  (0.0033) (0.010) (0.010) (0.016) (0.009) (0.009) (0.014)
Job change 0.0774*** 0.054*** 0.054*** 0.106*** −0.051*** −0.052*** 0.022
  (0.0038) (0.013) (0.013) (0.018) (0.012) (0.012) (0.016)
Region of birth - - Yes Yes - Yes Yes
Occupation Yes - - Yes - - Yes
Number of observations 4,655,054 223,213 300,290
Number of individuals 693,242 33,682 46,057
lnsig2u −0.4198 −12.415*** −12.416*** −0.493*** −14.026*** −14.030*** −0.455***
  (0.0091) (2.327) (2.327) (0.045) (3.824) (3.820) (0.040)
  Random effects probit, 2002–2008 – Wooldridge estimator with Mundlak correction
  Natives Western immigrants Non-Western immigrants
  III I II III I II III
Overeducation t-1 2.253*** 2.555*** 2.555*** 2.285*** 2.739*** 2.739*** 2.428***
  (0.004) (0.019) (0.019) (0.020) (0.019) (0.019) (0.018)
Overeducation t = 0 1.148*** 0.863*** 0.863*** 1.376*** 1.003*** 1.005*** 1.383***
  (0.007) (0.032) (0.032) (0.037) (0.034) (0.034) (0.034)
Job change 0.108*** 0.073*** 0.073*** 0.138*** −0.037*** −0.038*** 0.049***
  (0.004) (0.015) (0.015) (0.019) (0.013) (0.013) (0.017)
Region of birth - - Yes Yes - Yes Yes
Occupation Yes - - Yes - - Yes
Means over time Yes Yes Yes Yes Yes Yes Yes
Number of observations 4,655,054 223,392 300,712
Number of individuals 693,242 33,685 46,061
lnsig2u −0.155*** −1.211*** −1.212*** 0.038 −1.151*** −1.156*** −0.007
  (0.008) (0.067) (0.067) (0.036) (0.065) (0.065) (0.033)
  1. Note: The random effects probit models also include controls for: age, age squared, number of children, marriage, years of schooling and years since migration for immigrants, sector (five categories), county (21 categories), birth region for immigrants (six categories), and occupation (nine categories).
  2. The random effects probit models using the Wooldridge estimator and Mundlak corrections include means over time for age, age squared, number of children, years of schooling and years since migration. Standard errors in parentheses. ***p < 0.01, **p < 0.05, *p < 0.1.