Religion and family values in Europe: internal secularisation or reformation

1
Religion and family values in Europe internal
secularisation or reformation?

• Aart C. Liefbroer
• Arieke J. Rijken

2
Acknowledgement

• This paper is prepared within the project The
  Timing of Life, funded by NWO and the ESF
  Eurocores Programme on Human Values, Institutions
  and Behaviour (ESF HumVIB).

3
Background

• Christianity (like most religions) strongly
  supports marriage.
• Thus, religious people are found to hold
  unfavourable attitudes about behaviours that
  undermine the importance of marriage, such as
  unmarried cohabitation and divorce.
• However, it could be that the difference in
  family values between religious and non-religious
  differs by region.
• Opposing views could be put forward.

4
Internal secularization

• Religious people living in a secularized region
  could be subject to a process of internal
  secularization.
• Internal secularization means that religious
  doctrine looses its grip on the opinions of
  church members and/or that the translation step
  from doctrines to proper behaviour is left to the
  individual believer.
• If so, differences between religious and
  non-religious people in their family values are
  larger in regions where general levels of
  religiosity are high than in regions where
  general levels of religiosity are low.

5
Internal secularization
6
Reformation

• Religious people living in a low religiousness
  region may become a selective group that stresses
  traditional doctrines.
• Religious groupings may also emphasis traditional
  teachings in order to distinguish themselves from
  non-believers.
• If so, differences between religious and
  non-religious people in their family values are
  smaller in regions where general levels of
  religiosity are high than in regions where
  general levels of religiosity are low.

7
Reformation
8
Adaptation

• Levels of religiousness in a region may also
  impact non-believers.
• If non-believers adapt their values if
  religiousness is high in a society, the internal
  secularization hypothesis can be generalized, and
  it can be expected that the differences between
  religious and non-religious people in their
  family values are smaller in regions where
  general levels of religiosity are high or low
  than in regions where general levels of
  religiosity are intermediate.

9
Adaptation
10
Polarization

• But the reformation argument may also be
  generalized.
• Non-believers in highly religious regions may
  constitute a selective group, strongly differing
  in their opinions from the believers.
• If so, the differences between religious and
  non-religious people in their family values are
  larger in regions where general levels of
  religiosity are high or low than in regions where
  general levels of religiosity are intermediate.

11
Polarization
12
A cross-regional comparison

• Cross-national and cross-regional comparisons are
  useful, as religiosity varies across regions and
  countries.
• Focus on the regional level, for theoretical and
  practical reasons
• Theoretical Religiousness can vary within
  countries, and it could be that one is influenced
  by religiousness in ones region rather than in a
  country as a whole.
• Practical only 23 countries. Of course, regions
  are nested within countries. If the regional
  level becomes very detailed, the data are spread
  too thinly.
• Data from the ESS Wave 2006 can be used.

13
Relevant items

• How much do you approve or disapprove if a woman
  (man) lives with a partner without being married
  to him (her)?
• How much do you approve or disapprove if a woman
  (man) has a child with a partner she (he) lives
  with but is not married to?
• How much do you approve or disapprove if a woman
  (man) gets divorced if she (he) has children aged
  under 12?
• Answer categories strongly approve, approve,
  neither approve nor disapprove, disapprove,
  strongly disapprove

14
Measurement issues

• Marriage centrality is measured by the mean
  disapproval score on the three item unmarried
  cohabitation, parent while cohabiting, and
  divorce if young children. Higher score means
  stronger disapproval of these behaviours.
• Religious involvement is measured by a factor
  score based on three items religious
  self-evaluation, frequency of church
  attendance, and frequency of prayer. One clear
  factor higher score implies more religious
  involvement.

15
Disapproval by membership
16
Disapproval by membership
17
Level of religiousness across Europe
18
Modeling issues

• Three-level models, with individuals, regions and
  countries as levels.
• More than 37,000 respondents, 159 regions and 23
  countries.
• Start from a simple random intercept model,
  followed by random slope models in which the
  slope of religious involvement is allowed to vary
  across regions and countries.
• Next, interactions between individual and
  regional levels of religious involvement are
  added.

19
Results
20
Results
21
Influence of religiousness by regional
religiousness levels
22
Conclusions

• Family values vary more at individual and
  national level than at regional level.
• Effect of individual religious involvement on
  family values varies about equally at the
  regional and national level.
• Support for internal secularization hypothesis.
• Analysis at the country level confirms findings,
  but country-level variable and cross-level
  interaction is not statistically significant due
  to small number of cases.
• Next step finer-grained regional
  differentiation.
• Consider the role of selective migration.