Children and FAMILIES: Time Management and Perception OF STRESS

1. Children and FAMILIES:Time Management and Perception OF STRESS Elsa Fontainha ISEG – Technical University of Lisbon – Portugal 3rd International Conference International Society for Child Indicators University of York, 27th-29th July 2011

2. Aim of the Research Keywords: stress; subjective well-being; time use; intra-household data The aim of this paper is to shed more light on the subjective well-being of Portuguese children and parents using a unique micro database (Portuguese Time Use Survey, Office of National Statistics, Eurostat guidelines, N=8,389. Includes information about time allocation by activities and time management, perceived stress and preferred time allocation)

3. Children Well-Being and Stress • Stress can diminish subjective well-being (Ng et al., 2009:258). • Child well-being in Europe: two of the top ten indicators are subjective indicators (Bradshaw and Richardson, 2009:339). • Parental Time and Parental Stress… • Parenthood & time (demography, fertility, divorce) • Parental time & stress (direct effect?) • Parental time & other time trade-off (mother work mother absent?) • Parental time & market work (“babies or bosses”?)

4. Micro Data- Time Use Survey Portuguese TUS 1999, Eurostat Guidelines • Diary + Individual Questionnaire (*) + Household Questionnaire. • (*) time management, perceived stress and preferred time allocation Our research: • All sample N=8,389;Subsamples: 3 members hh • mother + father + child < 15 (0-5yrs old; 6-14yrs old); • mother + father + child 15+; child and parental data are combined

5. Stress Questions (examples ; total 16 questions; same as GSS Canada 1998)

6. Empirical Strategy • Descriptives (by family type) • Logit and Probitmodels (All; Parents) • Dependent variable: stress (1,0) • =1 (“Usually do you feel rushed?” Yes. Always and Often) • =0 all other cases

7. Families (#3) in a Rush I • Parents never or rarely had time to do whatever they want to do • (Mo 90%-Fa 84%;Mo 91%-Fa 83%) • Stress affects more mothers than fathers • Stress affects more the employed mothers than the non-employed mother

8. Families (#3) in a Rush 2 • Employed mothers’ feeling of lack of time in weekends (frequently + sometimes) is higher • (Mo 90%; Fa 29%) • time pressure in professional life is higher concerning fathers than mothers • children age seems not to have a large effect on fathers’ stress however is stronger among the mothers of older children (6-14 years old)

9. Is Stress an Heritage? In the each family children and parents data were combined. • The child feelings of being in a hurry are less frequent than the one felt by parents, but the gender difference is present in both groups. • Stress from parents to children(+15) (an heritage?) similar attitudes concerning slowing down in the future and gender differences

10. Feelings of Time Crunch Parents and Children (Family 3, Fa+Mo+C) Units: %

11. Models: Stress predictors • Logitand Probit models • Dependent variable: stress • Explanatory variables (predictors) : children number and age, gender, age, educational level, time of work, employment status, personal time, family size, etc. • Several specifications ( example >>>)

12. Empirical Results Stress Models *** p< 0.001 ** p< 0.005 * p<0.010

13. Models Results I • The signs of the coefficients (odd ratios) indicate • A positive association between employment status and education. Considering all the population and only the family adopted as reference – couple with one child – in general only 2 phenomena, education and total time of work predict well about 70% of the occurrences • A negative association at the aggregate level for age • The signs are the same for mothers and fathers • There are differences by gender • Children age (less stress related with less 6 years old)

14. Models Results • Total time of work, paid and unpaid is one of the fundamental stressors. Alternatively, short personal care time induces stress. Social time reduces it. • As the models are non linear (probit, logit), the coefficients are not constant and the effects of a change in an independent variable depend on the point at which that change occurs (next graphs illustrate)

15. 1,00 0,80 Stress probability 0,60 0,40 1 2 3 4 5 6 7 8 9 10 11 12 Educational level Stress – Predictions - Education

16. 1,00 0,80 Stress probability 0,60 0,40 1 2 3 4 5 6 7 8 9 10 11 12 Educational level Stress – Predictions - Education Changing the educational level and assuming the other variables equal to the mean values. > Level 8 = > 90% probability of stress

17. Stress – PredictionsTime of Work and Personal time Source: Model 5, Logit, N=653

18. Stress – PredictionsTime of Work and Personal time If the personal time (includes sleeping)increases from 9 hours a day to 10 hs, the predicted stress decreases from 81% to 78% Source: Model 5, Logit, N=653 (Mothers and Fathers 3 members families)

19. Conclusions I • Perception of stress is affected by gender, education, employment status, working time and personal time, family characteristics and child age. • There are signs that stress is widespread, is growing and is gaining ground among the new generations. It may be being passed from parents to children who live in the same household unit.

20. Conclusions II • Children less than 6 years old have no effects on or contribute to reduce stress perceived by parents. By contrast children over 6 years old seem to have some influence on stress, in particular mothers' stress. • Personal time and working time have inverse effects on predicted stress

21. Conclusions III • If the parents had more time, the preferred allocation for it is socializing with family (20%). But if the children (+15 yrs) had more time, the preferred allocation for it is socializing with friends (20%). These results converge with Bradshaw and Richardson (2009:344-345). • Entertainment, touristy travel and sports are activities equally preferred by both parents and children.