TURKS, SOMALIS AND DANES
CIVIL STATUS AFTER HOME-LEAVING
The analyses above supported hypothesis H2 in that marriage led to home-leaving for Turks, both for descendants and immigrants. This indicates that Turks do indeed choose the traditional path of marriage prior to home-leaving. To explore this further, competing risk models were estimated for the three outcomes of leaving home to live alone, to marry or to cohabit with a partner.
Table 5: Routes out of the parental home: numbers, distributions and selected hazard ratios.Table 5: Routes out of the parental home: numbers, distributions and selected hazard ratios.
- = reference category. **** p < 0.0001. N=115,937
The covariates from the previous models were all included but are not presented here.
Initially the different ethnic distributions for the routes out of the parental home should be noted. Less than 1% of Danes and Somalis moved directly into marriage. For Somalis, this means that only five people in the population took such a step. In con-trast, 27.3% of Turkish descendants and 44.9% of Turkish immigrants moved into marriage. Of the Danes, 29.6% moved directly into cohabitation with a partner while only 4.5-6.5% from the other three groups did so. The vast majority of Somalis moved into singlehood after leaving home. For Danes and Turkish descendants, 69.7% and 66.7% respectively lived as singles after leaving home and 50.6% of the Turkish im-migrants did so. Consequently, H2 is supported: Turks more often than Danes and Somalis left home for marriage. Turkish descendants seemed to some extent to have adopted Danish norms as the share moving into marriage was markedly lower than for Turkish immigrants. The HRs, however, told a different story as the HR for marriage was higher for Turkish descendants than for Turkish immigrants. This lends support for H3Alt that Turkish descendants are similar to Turkish immigrants in their pattern for leaving home.
To understand the differences better, the Cox regression models for Turkish immi-grants and Turkish descendants were examined in greater detail. The Cox models indicated interesting tentative explanations for the outcome differences between the two groups.
THE INDIVIDUAL
Gender differences were similar for Turkish immigrants and descendants: women were approx. 4 times as likely as men to leave home for marriage. This indicates more traditional home-leaving norms for female home-leavers than for male. The cohort indicated changed patterns over time with the older cohort having a higher hazard for
leaving home for marriage and, for Turkish immigrants, a lower hazard for leaving for singlehood. Home-leaving patterns thus seem to have become more similar to Danish home-leaving patterns. This could indicate straight-line assimilation taking place. With respect to income, the analyses showed that the higher the individual in-come, the higher the hazard for leaving home for marriage. Educational attainment also influenced the traditional pattern of home-leaving into marriage. Both immigrants and descendants with a higher level of education were more likely to leave home to live alone. Descendants with a higher level of education were less likely to move into marriage. Education thus leads to less traditional home-leaving patterns. In line with this, those who were studying had a slightly higher hazard for leaving home to live as a single and, for Turkish immigrants, a lower hazard to leave for marriage.
THE PARENTAL FAMILY
The result for the covariates on the parental family indicated the importance of tradi-tional norms in the parental family for the home-leaving patterns of the young adult.
Living in a large household led to a higher hazard for leaving for marriage than if living in a smaller household. Living with only one parent heightened the hazard for moving into cohabitation with a partner for both immigrants and descendants. Living with a father led to a higher hazard for living alone after home-leaving than if living with both parents. If the parents had divorced, a possible explanation could be that the break-down of the parental marriage leads to less traditional family values for the children. It could also be a consequence of single parents being less able to exercise pressure to uphold traditional values. Household income had a major effect for im-migrants in particular, with a higher household income leading to a lower hazard for moving into marriage and higher hazard for moving into cohabitation. For descend-ants, the pattern was similar for all outcome types with a higher household income leading to a lower hazard for leaving home but with a particularly low hazard to leave for marriage. It thus seems that the higher the financial prosperity of the parental fam-ily, the less traditional patterns regarding home-leaving and marriage. For immigrants, if the household head was anything else than in employment, the hazard to leave for singlehood was higher. Otherwise, the employment situation of the parental family showed no clear or significant pattern.
THE PARENTAL HOUSING UNIT
Less living space in the parental family, which again could be seen as an indicator for living in a more traditional family, resulted in a higher hazard for leaving home for marriage and, for Turkish immigrants, a lower hazard for singlehood and cohabitation.
The tenure of the parental home showed no clear pattern except the general tenden-cy of owner-occupation leading to a lower hazard for home-leaving. Finally, living outside of Copenhagen had a substantial effect on the hazard for leaving home to live
alone. While all the HRs indicate quicker home-leaving if living outside Copenhagen, they are not all significant. For leaving home to live alone, the HRs are significant and higher than for the other routes. This could indicate an acceptance of having to fore-go traditional norms in order to attend education, making it acceptable to live alone if studying. This is corroborated by the higher hazard of those studying for moving to singlehood.
Table 6: Results from separate competing risk models for Turkish immigrants and descendantsTable 6: Results from separate competing risk models for Turkish immigrants and descendants
Variables Turkish immigrants Turkish descendants
Single Married Cohabit. Single Married Cohabit.
Individual
x Man ‐ ‐ ‐ ‐ ‐ ‐
x Woman 1.183*** 4.360**** 1.384* 1.209**** 3.770**** 1.215
x Cohort 1980‐1988 ‐ ‐ ‐ ‐ ‐ ‐
x Cohort 1974‐1979 1.085 1.789**** 0.788 1.120* 1.382**** 0.801
x Cohort 1968‐1973 0.810*** 2.076*** 0.733 NR NR NR
z(t) Basic schooling ‐ ‐ ‐ ‐ ‐ ‐
z(t) Vocational training 0.964 1.014 1.041 0.960 0.959 0.962
z(t) Upper secondary (‘Gymnasium’) 1.363**** 1.008 1.510^ 1.576**** 0.663**** 0.939 z(t) Further education 1.784**** 1.457* 1.397 1.769**** 0.517*** 1.022
z(t) In employment ‐ ‐ ‐ ‐ ‐ ‐
z(t) Retired incl. disability 0.753 0.317* 0.000 0.745 0.063** 0.000 z(t) Unemployed or outside work force 1.010 0.876* 1.129 1.124^ 0.975 0.912
z(t) Studying 1.145** 0.681*** 0.696 1.118* 1.069 0.850
z(t) Income: up to approx. EUR 5400 ‐ ‐ ‐ ‐ ‐ ‐
z(t) Income: approx. EUR 5400‐10,700 1.679**** 1.825**** 1.839** 1.738**** 1.737**** 1.396 z(t) Income: approx. EUR 10,700‐16,100 1.110 3.193**** 1.260 1.278** 3.009**** 1.481 z(t) Income: above approx. EUR 16,100 0.786*** 4.472**** 0.868 1.218* 4.391**** 1.502
Parental family
z(t) 2‐5 people in household ‐ ‐ ‐ ‐ ‐ ‐
z(t) More than 5 people in household 0.873* 1.584**** 0.938 1.001 1.388**** 0.802
z(t) Live with both parents ‐ ‐ ‐ ‐ ‐ ‐
z(t) Live with mother 1.061 0.930 1.810** 1.004 0.731** 1.463^
z(t) Live with father 1.214** 1.107 1.776** 1.234^ 1.016 2.661***
z(t) HH: employed ‐ ‐ ‐ ‐ ‐ ‐
z(t) HH: Unemployed or outside work force 1.298**** 1.004 1.122 1.182* 0.897 1.222
z(t) HH: studying 1.180* 0.955 0.901 1.107 1.181 1.182
z(t) HH: retired 2.170** 1.187 0.000 1.338 0.917 3.286
z(t) HH income: up to approx. EUR 46,900 ‐ ‐ ‐ ‐ ‐ ‐
z(t) HH income: approx. EUR 46,900‐67,000 1.102^ 0.485**** 1.365 0. 958 0.452**** 0.762 z(t) HH income: approx. EUR 67,000‐87,100 0.996 0.263**** 1.577* 0.695**** 0.310**** 0.827 z(t) HH income: above approx. EUR 87,100 0.851^ 0.174**** 1.530^ 0.610**** 0.203**** 0.953
Parental home
z(t) At least one room per person ‐ ‐ ‐ ‐ ‐ ‐
z(t) Less than one room per person 0.849** 1.335*** 0.671* 0.927 1.858**** 0.946
z(t) Public housing ‐ ‐ ‐ ‐ ‐ ‐
z(t) Co‐op 1.083 1.266* 1.557 1.229^ 0.694^ 1.464
z(t) Owner‐occupied 0.801** 0.753*** 0.833 0.677*** 0.830^ 0.943
z(t) Private rental 1.002 1.182* 0.787 1.068 0.940 1.474
z(t) Copenhagen ‐ ‐ ‐ ‐ ‐ ‐
z(t) Three biggest cities outside CPH 1.824**** 1.120 1.247 1.641**** 1.322** 1.374 z(t) All other 1.444**** 1.250*** 1.112 1.351**** 1.194* 1.365^
N 2,190 1,944 194 2,342 960 211
^ p < .1 *p < 0.05 **p < 0.01 *** p < 0.001 **** p < 0.0001
‐ = reference category; x = time‐independent variables; z(t) = time‐dependent variables. NR = Not relevant. N=7,841
^ p < .1 *p < 0.05 **p < 0.01 *** p < 0.001 **** p < 0.0001
- = reference category; x = time-independent variables; z(t) = time-dependent variables.
NR = Not relevant. N=7,841
CONCLUSION
The purpose of this paper was to gain empirical insights into the ethnic differences in home-leaving patterns. Despite extensive research on native home-leaving patterns, ethnic differences are an under-researched area. Life course analysis as a theoretical perspective framed the analyses and led to the identification of the key covariates. The Danish registers were utilised for event history analysis, a tool highly suitable for life course analysis. Cox regression models were estimated for the event of leaving home and for the three outcomes of moving out to live as a single, to marry and to cohabit with a partner. Four hypotheses were tested.
The first hypothesis stated that differences in age at home-leaving would persist when controlling for covariates thus indicating cultural differences in home-leaving patterns.
While some differences disappeared when controlling for covariates, others persisted, thus supporting the notion of cultural differences in home-leaving patterns. The sec-ond hypothesis focused on union formation, stating that for Turks there would be a strong link between the two life events of marriage and home-leaving. A strong link between home-leaving and marriage was substantiated for Turks, but not for Somalis.
While there was evidence of traditional norms persisting, there was however also in-dications of change taking place towards less traditional patterns. Thus it seems that the Danish home-leaving norms and the second demographic transition have affected Turkish home-leaving patterns to some extent. This indicates straight-line assimilation taking place for Turks but at a slow pace. Somalis were in general much more simi-lar to Danes with the exception, despite the fact that Somalis as a group has lived in Denmark for fewer years than Turks. This suggests that Somali home-leaving norms are more similar to Danish norms or that Somalis have been more influenced by the Danish home-leaving norms despite a shorter group history in Denmark.
The two last hypotheses were the two opposing hypotheses of Turkish descendants being similar to Danes in their pattern for leaving home and of Turkish descendants being similar to Turkish immigrants in their pattern for leaving home. The empiri-cal evidence showed a middle-way. When controlling for covariates, the differences between immigrants and descendants disappeared almost completely. As such, this would indicate limited straight-line assimilation taking place between immigrant and descendant home-leavers. However, to the extent that the socio-demographic char-acteristics of descendants differ from that of their ancestors, home-leaving patterns become increasingly similar to those of Danes. In this way, straight-line assimilation is taking place.
All in all, while cultural patterns persist between the ethnic groups, there are indica-tions of assimilation taking place through concurrent normative and socio-economic assimilation. While this process seems to be slow moving for Turks, change is hap-pening. Consequently, ethnic differences in home-leaving patterns can be expected to diminish further over time.
ACKNOWLEDGEMENTS
Financial support from NORFACE research programme on Migration in Europe – Social, Economic, Cultural and Policy Dynamics is acknowledged.
APPENDIX
Figure A1: Comparison of survival rates, years since the year of turning 18 Figure A1: Comparison of survival rates, years since the year of turning 18
0,00 0,10 0,20 0,30 0,40 0,50 0,60 0,70 0,80 0,90 1,00
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Somali immigrants Turkish immigrant Turkish descendant Danes
Middle East & Northern Africa The rest of Africa
The rest of Asia
REFERENCES