Fig. 1

Predicted probabilities of being working conditioned on the number of (censored/uncensored) informal caregiving hours. Upper figure shows the predicted probability of being working after the estimation of a progit model for this outcome using as explanatory variables informal caregiving hours (censored at 16 h/day), sex, education and size of municipality of residence. Lower figure shows the predicted probability of being working after the estimation of a probit model for this outcome using as explanatory variables informal caregiving hours (maximum 24 h/day), sex, education and size of municipality of residence. Vertical bars denote confidence intervals at 95% significance