1. Parental divorce and family formation: a comparison of 10 European countries Montse Solsona Carles Simó Eighth Meeting of the European Network for the Sociological and Demographic Studies of Divorce Valencia, 14-16 October, 2010 Jeroen Spijker * Research project: “Family Trajectories after divorce. Gender, kinship and territory”. Funded by the Spanish Ministry of Science and Innovation (I+D+I Project Ref. CSO2008-00654)

2. Background Second demographic transition: o Shifts in values related to family life and children have weakened the ‘traditional’ nuclear family in favour of cohabitation, postponement of both the timing of marriage and child bearing, increase in divorce, etc. o Caused mainly by increases in female education, female labour force participation, …., efficient methods of contraception. o But:  Not always values  behaviour  Self-fulfillment (wealthy & high educated) vs. structural constraints (poor)  Regional differences in SDT patterns due to nature of family systems  Inconsistencies in this “transition”, e.g.  SDT phenomena not new;  Nordic countries, high in SDT values but also in fertility; and  southern European countries lower on SDT values but also low in fertility

3. Background Total divorce rate, 1980-2 to 2005-7 (3 year moving average)

4. Background Second demographic transition: o Shifts in values related to family life and children have weakened the ‘traditional’ nuclear family in favour of cohabitation, postponement of both the timing of marriage and child bearing, increase in divorce, etc. o Caused mainly by increases in female education, female labour force participation, …., efficient methods of contraception. o But:  Not always values  behaviour  Self-fulfillment (wealthy & high educated) vs. structural constraints (poor)  Regional differences in SDT patterns due to nature of family systems  Inconsistencies in this “transition”, e.g.  SDT phenomena not new;  Nordic countries, high in SDT values but also in fertility; and  southern European countries lower on SDT values but also low in fertility

5. Background As a consequence of increasing rates of the dissolution of cohabiting unions and marriages, new living arrangements and family forms of divorced and separated individuals have become more diverse and widespread (different effects of divorce legislation on repartnering and reparenting discussed in paper). However, despite an international convergence in family law and attitudes towards divorce in Europe, country differences in post-divorce repartnering and parenthood still exist today. Changes in divorce legislation increasing divorce (and post-divorce events): o Legalisation (permanent effect) o No fault (permanent effect) o Unilateral (temporary effect)

6. to analyse post- first-union trajectories in Europe repartnering reparenting Objective

7. Data issues and methodology Data: Fertility and Family Survey (FFS), supported by data from Eurostat (New Cronos database) and National Statistical Institutes Advantages of using FFS for our study: It allows separating the moment of union dissolution from events that occur afterwards, e.g. survival functions and probabilities of transitions to new relationships and parent- hood according to different types of post-divorce living arrangements can be estimated. Drawbacks of using FFS in our study: Only adults until the age of about 50 were interviewed More women than men were included (in some countries no men were interviewed) Country-specific sample sizes are reduced substantially if only respondents who have experienced a union dissolution are considered As most of the surveys were conducted in the 1990s, results do not fully reflect actual patterns of post-dissolution living arrangements. Methodology: The FFS was used to estimate survival functions of entering into a new partnership and new parenthood after experiencing one’s first union dissolution, as well as the odds for legally divorced and separated respondents of entering into a new partnership. The other 2 sources were used for the calculation of divorce and remarriage indicators.

8. Sample 1Type of 1 st union (%)Post-1 st union (%)Sample 2 1+ union break-upsmarriageconsensual repartner- ing (re)paren- ting divorced/ separated only Country, year Total ♂♀ ♂♀ ♂♀ ♂♀ ♂♀♂♀ Austria, 1995-96 1255309946253675645747282994414 Belgium, 1991-92 5742393356367373356534232124191 Czech Republic, 1997* 434109325465554458363364227159 Estonia, 1994 78926652328 72 70673231113#209# Finland, 1989-90 12513628892639746168582934133472 France, 1994 156952610432540756065512629211528 Germany, 1992 164156210792537756359533639207544 Hungary, 1992-93 8322655675769433161593328118289 Norway, 1988-89 10622957673932616866563841111282 Slovenia, 1994-1995 25782175496151396365142049109 Spain, 1994-95 322106216356165394541212734117 Total Sample 998631216865344366576355303211213314 Data issues and methodology FFS sample characteristics of respondents experiencing a union break-up * The Czech Republic was not analysed because male respondents who were not in a relationship at the time of the survey were not interviewed; # Estonia was not used in the analysis that only contained divorced and separated respondents as just one divorced/separated woman lived with a new partner

9. RepartneringNew parenthood 1. Differences or not between sexes Men enter faster in new partnership or parenthood after end of 1 st union than women Detailed results shown here 2. Differences or not between 1st union type More likely to occur if 1 st union was consensual rather than marriage Detailed results shown here 3. Differences or not between age groups Younger cohorts enter faster in new partnership or parenthood after 1 st union than 35+ pop 4. Differences or not between previous fertility Those without children enter faster in new partnership or parenthood after end of 1 st union than those who are already parents Differences in repartnering and new parenthood according to sex, 1 st union type, age and previous fertility. Overview of analyses performed and examples shown in presentation. Hypotheses

10. Males Females Males – censured Females – censured Time from the end of the 1 st to the beginning of the 2 nd union (months) France, 1994* Survival functions from the end of the 1 st union to the beginning of a 2 nd by sex Results Accumulated survival Austria, 1995-96* Finland, 1989-90* Spain,1994-95* Norway 1988-89 Germany, 1992* Slovenia, 1994-96Hungary, 1992-93Estonia, 1994 Belgium 1991-92 * Sex differences statistically significant (p<0,1) according to the Mantel-Cox Log Rank Equality Test

11. Marriage Consensual union Marriage – censured Consensual union – censured Time from the end of the 1 st union to (new) parenthood (months) France, 1994* Survival functions from the end of the 1 st union to new parenthood by union type Results Accumulated survival Austria, 1995-96*Finland, 1989-90* Spain,1994-95* Norway 1988-89* Germany, 1992* Slovenia, 1994-96Hungary, 1992-93Estonia, 1994 Belgium 1991-92* * Union type differences statistically significant (p<0,1) according to the Mantel-Cox Log Rank Equality Test

12. Tentative explanations Variable 2 nd unionPost 1 st union parenthood bp-value B Median time (months) from end of 1st to beginning of 2nd union Nottested2,2780,004 Youth unemployment (%) 0,9160,011- Ln of GDP 14,2050,069- Female labour force activity (%) -1,6710,029 Post-secondary completed (% F 25+) -5,6560,058 Legalisation of divorce (yr) -- Female share of part-timers (%) -- Constant -101,3280,159-125,6730,096 N 10 Adjusted R 2 0,5500,667 Durbin-Watson 1,6502,603 Results of regression analysis *of the length of time (months) between the end of the 1 st union to the beginning of the 2 nd and new parenthood. Women only. * Stepwise regression method applied where only significant (p<0,1) variables were included in the final models.

13. Summary Value changes and economic uncertainty have led more people to separate or divorce and for less traditional forms of family formation to become more common Our results for women showed three different patterns of post- divorce trajectories: Both new unions and post-first-union parenthood occur relatively quickly (Norway, Germany, Hungary, Belgium); New unions may take place as quick as in the first group but the transition to new parenthood is slower (Austria Slovenia, Finland, France, and Estonia); Takes a long time for both transitions to occur (Spain). What was not shown in presentation but can be found in paper Country and time differences in the total divorce rate Remarriage rates (remarriage is not always the preferred repartnering option, even among the legally divorced).

14. Summary Median duration to repartnering vs. reparenting after the dissolution of the first union.

15. Summary Part of the international differences in repartnering and reparenting can be accredited to differences in divorce legislation (González, Viitanen, 2009), but as we were able to show with FFS data: being male, under 25, having had a consensual union as first union; and a childless first union were often associated with both post-first union trajectories in the countries that were analysed: Especially in Austria, Germany, France, Finland and Spain (traditional pattern of repartnering and post first-union parenthood?) Few differences in Estonia, Slovenia, Hungary and Belgium (a more uniform pattern). Norway - somewhere in the middle.

16. Summary RepartneringNew parenthood 1. Differences or not between sexes Men enter faster in new partnership or parenthood after end of 1 st union than women Austria, Finland, France, Germany, Spain Austria 2. Differences or not between 1st union type More likely to occur if 1 st union was consensual rather than marriage Austria, Finland, France, Germany Austria, Belgium, Finland, France, Germany, Norway, Spain 3. Differences or not between age groups Younger cohorts enter faster in new partnership or parenthood after 1 st union than 35+ pop Austria, Estonia, Finland, France, Germany, Hungary, Norway, Spain 4. Differences or not between previous fertility Those without children enter faster in new partnership or parenthood after end of 1 st union than those who are already parents Austria, Belgium, Finland, France, Germany, Hungary, Norway, Slovenia, Spain Differences in repartnering and new parenthood according to sex, 1 st union type, age and previous fertility. Summary of results. Source: FFS data (own calculations)

17. It would seem that it is either a lack of personal income security that delays union formation or general economic welfare: As the regression analysis showed: o Youth unemployment was associated with second union formation (e.g. low in Spain, unemployment rate > 40%, and high in Germany, unemployment rate 6%); o In Eastern European countries it was their low GDP rather than unemployment that was associated with a low median time in repartnering. This seems paradoxical, but accords with results from other studies: o e.g. Sobotka (2008) sees the former as a response to different structural conditions marked by economic crisis; o while delaying marriage and family formation is one strategy for improving one’s own welfare. Discussion

18. A copy of the full paper is available on request. For any questions or comments don’t hesitate to send me an email. (jjaspijker@yahoo.es) THANK YOU FOR YOUR ATTENTION !