| Popis: |
Marital stability is an important topic in studies on family because of its meaning for fertility and for the well-being of children and partners. The rise of the divorce rate observed in Poland since the second half of 1990s raises a question about factors correlated with the marital disruption risk. In relevant research one can distinguish between those factors which are related to partners’ characteristics, to features of union and to the context. The paper starts with some theoretical considerations on the divorce correlates. Some references to empirical findings regarding other countries are presented as well. Next, the event history regression of first marriages disruption is estimated. The model is specified as a piecewise constant exponential model with proportional relative risks. The data used comes from the Education, Family and Employment Survey from 2006. Most of the estimation results are consistent with findings for other countries: marriages with premarital children or contracted while expecting a child, marriages of women brought up in bigger cities and those of employed women were less stable. The change of the educational gradient of divorce is an important finding – in the studies pertaining to the period before the 1990s the women with higher education showed a higher risk of divorce. In the current study, which refers to the years since the mid-1980s to 2006, marriages of best educated women have the lowest risk of disruption. Surprisingly, marriages preceded by cohabitation do not have an elevated disruption risk compared to direct marriages. Children’s impact on the disruption risk is lower than expected – only marriages with very small children (0-2 years old) are more stable, parity and presence of older children does not make a difference as compared with couples without children. In conclusions, some suggestions have been formulated regarding the data sources for the future research on unions’ stability. |
