Causes of Religiousness: Parental Influences in Germany

by Mirko

1. Introduction

Religion plays an important role in present identity discourses in Europe, particularly in Western European countries like France, Belgium and Germany. In the latter case, recent studies have shown a decline in the importance of Christian majority denominations for individual identity formation.¹ To better understand these dynamics, one might ask what makes some individuals more religious than others.
Recently, a qualitative study of Salafist youth cultures in Germany has suggested that young people become deeply religious Muslims to rebel against their parents if they were not raised religiously during their upbringing.² This poses the question whether the relationship between the religiousness of young people and the role of religion in their families is applicable in a wider context. One might ask whether this form of rebellion is not only common in specific Islamic circles, but for young people in general who use their religiousness to set themselves apart from their parents: How does the importance of religion in the upbringing of young people in Germany affect their religiousness?

2. Analysis

This question is explored with the ALLBUS³ dataset from 2012, a biennially trend study on attitudes and social structures in Germany.⁴ The population to be observed consists of all individuals between 18 and 30 years who had their primary residence in Germany between April and September 2012 (survey period).⁵ The hypothesis is that the smaller the role of religion during their upbringing (RUP), the more religious young individuals see themselves.⁶

Figure 1 illustrates the relationship between religiousness and RUP with box plots. The distribution of religiousness is displayed for every value of RUP. Each blue box contains 50% of the responses. Inside each box, the black line demarcates the median (“middle value”) of all responses.

Overall, the box plots point towards a positive instead of the expected negative relationship: the more important religion was during upbringing, the more religious young people are. However, the working hypothesis does appear to be correct in assuming that the relationship is linear. Suspicious are the outliers in the “0”-category on the x-axis (in red), which we would have assumed to be the majority of cases, since they claim to be more religious despite saying that religion has played no role in their upbringing. Unfortunately, there are too few of these cases for a meaningful investigation.

Instead, we substantiate the first impression that the relationship between religiousness and RUP runs in the opposite direction of the initial hypothesis. Since the relationship does appear to be linear in Figure 1, this analysis uses a simple multiple linear regression analysis. To make sure that we do not confuse the effect of RUP with the effect of other variables, three controls that could influence both religiousness and RUP are included in the models. The first is the origin of the respondents, to check whether they spend their childhood in West- or East-Germany, or in another country. The former distinction is of particular importance, since East-Germans have been found to be less religious than West-Germans.⁷ The two other factors controlled for here include the political orientation of the respondents, as well as the highest professional education of the parents.⁸ Table 1 shows the regression estimates for these variables in interaction with the role of religion in upbringing:

Table 1. Determinants of Religiousness among Young People in Germany

DV: Religiousness	(1)	(2)	(3)	(4)	(5)
Religion in Upbringing (RUP)	0.58***	0.52***	0.58***	0.65***	0.63***
	(18.74)	(13.00)	(18.72)	(5.72)	(5.54)
Region: East Germany		-0.66*			-0.72*
		(-2.26)			(-2.46)
Region: Other		1.42			1.70*
		(1.67)			(2.01)
East Germany * RUP		0.17*			0.17*
		(2.21)			(2.15)
Other * RUP		-0.091			-0.19
		(-0.69)			(-1.44)
Political Orientation			0.045		0.035
			(0.70)		(0.54)
Apprenticeship				-0.19	0.10
				(-0.26)	(0.14)
Professional School				-0.25	-0.087
				(-0.34)	(-0.12)
Applied Sciences College				-0.44	-0.072
				(-0.51)	(-0.08)
College				-0.79	-0.61
				(-1.02)	(-0.79)
Apprenticeship * RUP				-0.14	-0.19
				(-1.12)	(-1.50)
Professional School * RUP				-0.17	-0.20
				(-1.36)	(-1.62)
Applied Sciences College * RUP				-0.10	-0.19
				(-0.66)	(-1.16)
College * RUP				0.11	0.069
				(0.80)	(0.53)
Constant	2.07***	2.34***	1.85***	2.42***	2.31**
	(14.40)	(11.76)	(5.31)	(3.51)	(3.15)
R-Squared	0.39	0.40	0.39	0.41	0.42
Observations	553	553	553	553	553

Note: OLS estimates with t-stats in parentheses. * p < 0.05, ** p < 0.01, *** p < 0.001
Source: ALLBUS

All models confirm the observations from Figure 1: There is a significant, positive relationship between the role that religion plays in upbringing, and the religiousness of young people in Germany. Model (5) in the table furthermore indicates that this effect is stronger for East-Germans compared to West-Germans.⁹ Political Orientation appears to have no significant relationship to religiousness. The same goes for the effect of religion in upbringing on the condition of a specific educational background: the interaction effects in model (5) are not significant at the 5% level, which is to say that we cannot make any inferences from this sample to the wider population.¹⁰
Measuring religiousness only based on the respondents self-perception may not give us an entirely reliable image of the “true” religiousness of individuals. We therefore look at two alternate measures for religiousness: church visits and an index of questions about the respondents beliefs (e.g. “I do not care if there is a god”). Tables 5 and 6 in the appendices show that the effect of RUP is significantly lower compared to table 2 for these measures. In case of church visits, we could argue that ritual religiousness is conceptually different from religious beliefs. In case of the sum score however, we have to consider that a self-assessment may be a less reliable measure for religiousness than originally assumed in our analysis.

3. Conclusion

It is not surprising to find that young people are greatly influenced by their parents. However, this analysis has some relevance in the context of the conclusions that were drawn for the specific youth culture of Salafism. It shows that religiousness as a form of protest among young people is not a wide-spread phenomenon. This phenomenon is likely to be limited to a specific sub-culture that uses religious symbols.
At the same time, the above analysis suffers from several weaknesses. First, due to a lack of cases, we were unable to explore the relatively smaller group of young people who considered themselves to be both very religious and of a non-religious family. And second, we are uncertain whether a self-assessment of religiousness captures the concept of religiousness that we want to measure. The same holds true for the primary independent variable of religiousness during the respondents upbringing.
A reasonable next step would be to construct a more profound theoretical concept of religiousness before applying statistical methods to it. Further research could also investigate the validity of these results for other generations, and, given additional data, look towards the religiousness of less common (non-Christian) religious denominations.

Appendices

Table 2. Summary Statistics

Variable	Mode	Mean	Std. Dev.	Min.	Max.	Description
Religiousness	1	4.03	2.98	1.0	10.0	Assumed interval scale, 1=Min, 10=Max
Religious Service Visits	1	2.09	1.23	1.0	9.0	Ordinal scale, ranging from never to daily
Religious Beliefs (Mean of Sumscore)	2.75	2.40	0.95	1.0	7.8	Based on 4 questions, each with 4-item Likert-Scale
Role of Religion in Upbringing	0	3.40	3.23	0.0	10.0	Assumed interval scale, 0=Min, 10=Max
Region	0	0.38	0.58	0.0	2.0	Region where R grew up, 0=West, 1=East, 2=Other
Political Orientation	5	5.02	1.56	1.0	10.0	Political Orientation of R, 1=Left, 10=Right
Education of Parents	3	3.06	1.20	1.0	5.0	Highest professional education of parents, assumed categorial
Observations	553

Table 3. Main Effects of Control Variables

DV: Religiousness	(1)	(2)	(3)
Religion in Upbringing (RUP)	0.56***	0.57***	0.55***
	(17.48)	(18.77)	(16.92)
Region: East Germany	-0.30		-0.29
	(-1.31)		(-1.24)
Region: Other	0.87*		0.77
	(2.01)		(1.76)
Apprenticeship		-0.85*	-0.74
		(-2.02)	(-1.75)
Professional School		-0.95*	-0.86*
		(-2.27)	(-2.06)
Applied Sciences College		-0.93	-0.77
		(-1.84)	(-1.50)
College		-0.51	-0.43
		(-1.18)	(-0.98)
Constant	2.19***	2.89***	2.91***
	(12.71)	(7.09)	(7.03)
R-Squared	0.41	0.41	0.41
Observations	575	575	575

Note: OLS estimates with t-stats in parentheses. * p < 0.05, ** p < 0.01, *** p < 0.001
Source: ALLBUS

Table 4. Simple Linear Regressions on Control Variables

DV: Religiousness	(1)	(2)	(3)
Region: East Germany	-1.65***
	(-6.06)
Region: Other	1.74**
	(3.12)
Political Orientation		0.058
		(0.72)
Apprenticeship			-1.91***
			(-3.42)
Professional School			-2.00***
			(-3.62)
Applied Sciences College			-1.77**
			(-2.63)
College			-0.88
			(-1.51)
Constant	4.40***	3.74***	5.64***
	(29.75)	(8.76)	(11.04)
R-Squared	0.09	0.00	0.04
Observations	553	553	553

Note: OLS estimates with t-stats in parentheses. * p < 0.05, ** p < 0.01, *** p < 0.001
Source: ALLBUS

Table 5. Determinants of Religiousness Measured with Service Visits

DV: Service Visits (Logged)	(1)	(2)	(3)	(4)	(5)
Religion in Upbringing (RUP)	0.11***	0.11***	0.11***	0.12***	0.12***
	(19.89)	(14.87)	(19.87)	(6.00)	(5.95)
Region: East Germany		-0.029			-0.035
		(-0.55)			(-0.66)
Region: Other		0.052			0.078
		(0.34)			(0.51)
East Germany * RUP		0.015			0.0086
		(1.04)			(0.60)
Other * RUP		-0.0092			-0.015
		(-0.39)			(-0.62)
Political Orientation			-0.0014		-0.0057
			(-0.13)		(-0.49)
Apprenticeship				0.16	0.19
				(1.23)	(1.39)
Professional School				0.19	0.21
				(1.45)	(1.57)
Applied Sciences College				0.20	0.23
				(1.33)	(1.49)
College				0.14	0.15
				(1.03)	(1.09)
Apprenticeship * RUP				-0.032	-0.037
				(-1.42)	(-1.59)
Professional School * RUP				-0.014	-0.018
				(-0.63)	(-0.80)
Applied Sciences College * RUP				-0.011	-0.018
				(-0.41)	(-0.61)
College * RUP				0.0100	0.0081
				(0.43)	(0.34)
Constant	0.21***	0.22***	0.22***	0.051	0.073
	(8.31)	(6.27)	(3.55)	(0.41)	(0.55)
R-Squared	0.42	0.42	0.42	0.44	0.44
Observations	553	553	553	553	553

Note: OLS estimates with t-stats in parentheses. * p < 0.05, ** p < 0.01, *** p < 0.001
Source: ALLBUS

Table 6. Determinants of Religiousness Measured with Beliefs

DV: Religiousness (Mean of Sumscore)	(1)	(2)	(3)	(4)	(5)
Religion in Upbringing (RUP)	0.16***	0.13***	0.16***	0.17***	0.15***
	(14.98)	(9.29)	(14.97)	(4.17)	(3.87)
Region: East Germany		-0.48***			-0.49***
		(-4.82)			(-4.82)
Region: Other		0.67*			0.69*
		(2.32)			(2.37)
East Germany * RUP		0.078**			0.077**
		(2.95)			(2.81)
Other * RUP		-0.059			-0.066
		(-1.31)			(-1.43)
Political Orientation			0.0099		0.014
			(0.45)		(0.61)
Apprenticeship				-0.069	0.13
				(-0.27)	(0.49)
Professional School				-0.026	0.078
				(-0.10)	(0.31)
Applied Sciences College				-0.091	0.13
				(-0.30)	(0.45)
College				0.063	0.18
				(0.23)	(0.67)
Apprenticeship * RUP				-0.020	-0.050
				(-0.43)	(-1.12)
Professional School * RUP				-0.0081	-0.024
				(-0.18)	(-0.56)
Applied Sciences College * RUP				0.0072	-0.031
				(0.13)	(-0.55)
College * RUP				-0.0067	-0.024
				(-0.14)	(-0.52)
Constant	1.87***	2.07***	1.82***	1.90***	1.89***
	(37.17)	(30.35)	(14.94)	(7.78)	(7.43)
R-Squared	0.29	0.33	0.29	0.29	0.34
Observations	554	554	554	554	554

Note: OLS estimates with t-stats in parentheses. * p < 0.05, ** p < 0.01, *** p < 0.001
Source: ALLBUS

1Pollack, Detlef/Müller, Olaf 2013: Religiosität und Zusammenhalt in Deutschland. Bertelsmann Stiftung, Gütersloh, online at: http://www.bertelsmann-stiftung.de/cps/rde/xbcr/SID-3237E893-FD93CE81/bst/xcms_bst_dms_37661__2.pdf, viewed 01.10.2013.

2El-Mafaalani, Aladin 2014: Salafismus als Jugendkulturelle Provokation. Zwischen dem Bedürfnis nach Abgrenzung und der Suche nach Habitueller Übereinstimmung, in: Schneiders, Thorsten G. (ed.): Salafismus in Deutschland. Ursprünge und Gefahren einer Islamisch-Fundamentalistischen Bewegung, Bielefeld: Transcript, 355–362.

3“General Population Survey of the Social Sciences”: http://www.gesis.org/allbus, translation by author.

4ALLBUS is (to my knowledge) the most comprehensive and recent social survey available for this country, and includes questions regarding both the religiousness of the respondents and the role religion played in their upbringing.

5The definition of “young” as individuals between 18-30 years old is chosen because it leaves enough time for parental influence to take place (adulthood), but does not go beyond the argument that the hypothesized phenomena relates to current youth cultures. The assumption is that older cohorts might show different behavior if the hypothesis proves to be true (see conclusion).

6Both religiousness and religion during upbringing are interval-scaled, ranging from 1-10 and 0-10, respectively. See the summary statistics in the appendices for more information.

7Pollack, Detlef/Müller, Olaf 2013: Religiosität und Zusammenhalt in Deutschland. Bertelsmann Stiftung, Gütersloh, online at: http://www.bertelsmann-stiftung.de/cps/rde/xbcr/SID-3237E893-FD93CE81/bst/xcms_bst_dms_37661__2.pdf, viewed 01.10.2013.

8For more information, see the summary statistics in Appendix I.

9See also Figure 2 in the appendices.

10Comparing these results to the main effects of region and education (see Table 3 in appendices) does indicate that religiousness is slightly lower among populations growing up in East Germany, and higher among those growing up outside the country, holding upbringing constant. This confirms previous results mentioned above.