Antonianum LXXXV (2010) 243-271
IS THERE A LIMIT TO RELIGIOUS DECLINE?
Social Secularization as the Main Context for heology in Europe
Lluis Oviedo, Antonianum University, Rome - Italy*1
Manuel Canteras, University of Murcia - Spain
Summary: heology needs to pay attention to the “signs of the time” as a requirement to cope with
current challenges. An analysis on the ongoing process of secularization in Europe, or religious
downward spiral, and its hypothetical limits is essential. he perceived negative consequences of
extreme secularization, and the cyclical nature of religious decline and renewal in the past may be
adduced as arguments in favour of such a limit. hese hypotheses, however, cannot easily be tested,
as secularization today has gone beyond former periods of religious drop, and several European
countries show no ending to that process. he analysis of available data sends mixed signals when
the main European countries are examined. An alternative approach consists of testing links between religious and social crises. Again the analysis does not provide strong evidence for the alleged
thesis, even if neither supports other current theories pointing to a progress of religious decline in
Europe. heology should review some dominant ideas regarding survival of religion while aiming
to restore credibility to its message.
Many theological voices claim that their endeavor needs to be better
framed in reference to the most pressing traits of the environment. he discussion even points to the intrinsic contextual character of every theological
program; in other words: theology would always be dependent – in one way
or another – on current ideas, cultural moods and social circumstances (Pears 2009; Schreiter 1985; Bergmann 2003; Bevans 1992). Yet the ongoing
process is somehow deluding; most of the suggestions have to do with indigenous beliefs and customs; with social or political issues; with feminism and
gender identities; and with dominant religions in their own environment.
Some suggestions point to postmodern cultural characteristics as a kind of “invisible” but nevertheless pervasive frame, inluencing in a subtle way our faith
assumptions.
More needs to be done to render efectively the program of a “contextual theology”. In our opinion, such a program requires an assessment of the
factors in the environment afecting theological production and Church life,
* Correspondence: loviedo@antonianum.eu
244
Lluis Oviedo - Manuel Canteras
both, in a positive way, inspiring new expressions; and in a negative one, in
the form of challenges to be tackled.
he characteristics of the designed program support the conviction that
theology would be made in a diferent way when moving from one continent to another, and even from one country to another, and even more, from
the city to rural areas. You cannot propose the same theological content to a
Western European audience, deeply afected by silent religious decline; and
an audience in USA, especially in some areas, where most people keep sound
religious beliefs and practice on a regular basis. It would pose diferent challenges when theology addresses a population with high levels of studies and
who are well informed about scientiic advances.
he thesis we propose in this paper is that, when taking a broader and
more demanding view on theology’s contexts, there is no alternative but to
consider the secularization process as the main challenge that theology confronts. here is a second idea which needs to be clearly stated: this is by no
means a “positive context” in the sense of ofering opportunities for a better
appreciation of our traditional faith. his is an old standpoint, which needs
to be clearly overcome in light of an evermore threatening state of afairs.
Secularization in Western Europe has reached such proportions, that a good
deal of the Christian message, the meaning of the Church and its mission
need to be rethought in light of ongoing events.
he present study will not do a complete job; it would be pretentious to
suggest so. What we intend is just to blow the whistle, to alert our colleagues
about the seriousness of the situation, and to invite them to react and to
build a theology more engaged in the urgent task of ofering answers to this
impeding process of religious decline.
he proposed analysis will attempt to be an exercise in reading the “signs of the time”, a theological imperative after the Second Vatican Council.
It will use in a technical way the available literature to review the ongoing
theories on this process, and – closely related – will examine in depth the
ideas regarding the “limits of the process of secularization”, which could be
assumed as a contrary view of the claim that one way or another, humans
need religion, and so we can conidently expect that this process will not
endure in the long term, and will not bring so much trouble, since the cycle
will – sooner or later – revert back to better times for institutional religion.
he following pages show how complex the panorama is in Europe at
the moment, when contrasting available data, and why theology cannot be
too conident, but should rather keep alert and apply itself to the urgent task
of coping with such a situation of dire emergency for the faith.
Is there a limit to religious decline?
245
1. Secularization Trends and its Limits
he scientiic study of secularization, carried out in recent years, has reaped beneits and at the same time has roused new perplexities. Some of the
fundamental trends have been assessed, and few can doubt the extent and
deepness of religious decline in most advanced societies. he main theories
available – classical, supply side, and demographic – have contributed to
explaining several of its many facets. Nevertheless other aspects of religious
decline remain obscure and are objects of hefty discussion, particularly where its future projections are concerned.
Religious decline in Western Europe, in the opinion of many sociologists, seems to be an unstoppable process that proceeds in a quasi-linear
way, as predicted by the classic “theory of secularization”. here is much
data that conirms this “traditional wisdom” (Bruce 2002, Voas and Crockett
2005, Voas 2009). Such a process, perhaps for the irst time in history, could
eventually lead to unreligious societies, not through constraint by political
or administrative pressures, as was the case under Communist rule, but in a
spontaneous way. However, from varying points of view, it is hard to imagine
societies without religion, or where the functions of religion are completely
absorbed by other social agencies. Such a diiculty justiies the idea of “intrinsic” or “natural” limits to the secularization process.
he case for the “limits of secularization” has been made several times
and from diferent perspectives. A quick review identiies the following theories:
• From an historical point of view, modern secularization has been seen
as a stage in a kind of cyclical movement including also revivals or moments of returning to religion. he historical record of the XIX and XX
centuries clearly shows how these cycles happened more than once after
the French Revolution, leaving the issue rather unclear (Rémond 1999;
Goldstein 2009).
• From the social theory standpoint, some sociologists have pointed out
the unavoidable nature of religion, at least in dealing with speciic social
needs and functions. According to them, secularization would deprive
society of some functions that are essential to its good performance and
the maintenance of order and cohesion (Luhmann 1989, 349 f ).
• From a more empirical point of view, sociologists have observed trends
in the last twenty years that lead to the conclusion that secularization
processes are receding in many settings. his data, therefore, bears evidence against the main theory regarding the progressive loss of religious
246
•
•
•
•
•
Lluis Oviedo - Manuel Canteras
indicators in modernized or modernizing societies (Stark 1999; Berger
1999; Jenkins 2007, 47 f., Greeley 2002).
From a more anthropological point of view, considering that human
beings feel religious or spiritual needs, it is not easy to imagine a society
without any kind of spiritual expression; its absence would represent a
loss for many programs of human growth and accomplishment (Berger
1974; Taylor 2007; Knobloch 2006).
he cognitive approach to religion furnishes a hypothetic model that
suggests the presence in the human mind of a kind of “attractor” or
gravity point, which draws ideas or feelings into a sort of religious representation, transforming some perceptions into subjective patterns of
transcendence or supernatural insight. Such an attractor could hardly
be completely suppressed without changing the nature of the human
mind (McCauley and Lawson 2002; Atran 2002).
Several scholars point to a transformation in the religious ield, suggesting that it changes form and expression, but does not disappear or fade
away. Perhaps it migrates from a traditional form, with more innovative
or competitive expressions taking the place of the declining ones (Stark
and Finke 2000).
Another empirical approach points out the negative consequences of
secularization processes in advanced societies: when the declining indicators of religious practice strongly correlate with signs of social and individual distress (increase of divorces, family crisis situations, and other
diverse pathologies) the limit of secularization becomes the limit of the
possibility of society itself (Giddens 1991).
Finally, some studies point to an absolute limit to the secularization
process, since it afects, negatively, fertility rates, to the point of endangering the future of secularized societies, giving place to more proliic
– and religious – populations (Kaufmann 2007, 2008).
However, other studies point to an inverse relationship between religiosity and patterns of “existential security”, welfare spending, and income inequality at cross-national level (Norris and Inglehart 2004; Gill and Lundsgaarde 2004; Paul 2005; Rees 2009). hese studies would demonstrate that,
being religion dependent on such patterns, its decline could be explained by
some indicators of social improvement. On the other hand, some voices have
been raised pointing to intrinsic weakness of such studies (Moreno-Riaño,
Smith and Mach 2006; Kaufmann 2008; Young 2009), since cross-national
Is there a limit to religious decline?
247
comparative investigation and some statistical applied methods resent of
several methodological problems. Furthermore Kaufmann has been very vehement on signaling the “demographic limit” of religious decline, an idea
already considered by Norris and Inglehart in a weaker way.
A diferent path can be tried. he perceived negative tendencies (divorces, falling fertility, educational troubles…) which could be associated with
religious crisis – both at the personal and social levels – can be assumed as
symptoms of a malaise consequence of the diminishing presence of transcendence in modern life. Nevertheless, an alternative view points towards a
theory of transition and new stability, which claims that Western societies go
through a hard and slow process of adjustment to new structural and cultural conditions, while they are still in the process of developing. herefore, as
is the case in any transitional process, manifest forms of resistance to change,
dis-adaptations, cultural dissonance, and other negative symptoms may be
expected. According to this theory, it is simply a question of time until the
process matures and is completed, wherein the troubles and negative tendencies arising in that phase of crisis can stabilize and reach new standards and
better social expressions. However, a diferent hypothesis could mature, considering the riskiest probability: that the process of secularization and social
crisis reaches a point of no return, with very catastrophic consequences, in
a similar vein to the most ominous predictions concerning climate change.
Before continuing, it is prudent to remind ourselves of some terminological issues. he concerns regarding the use of terms like “religion” and
“secularization” are well known (Moreno-Riaño, e.al. 2006). hese pages
will not address the thorny questions regarding their meaning and limits
of application, if any. It has become a convention to distinguish between
“secularization” – macro-social and institutional processes – and “religious
decline” – rather at individual level (Casanova 1994, Dobbelaere 2002, Turina 2007). Our research deals with “secularization” understood as the decline in the empirical indicators of religious practice and self-description of
individuals. We leave aside other interesting questions, for example those
regarding the structural dimension of that trend, and its broadest cultural
and historical aspects.
he hypothesis about the limits or threshold of secularization ofers an
opportunity to review the state of religions decline and its trends. his idea
is deeply entrenched with several other theories trying to understand the
ongoing religious crisis. To test the hypothesis we will, irst, resort to a sociometric study, using available European series of data. A second step will
be to review current relevant theories on religious decline and its causes in
248
Lluis Oviedo - Manuel Canteras
light of the statistical evidence. A case study will deliver some proof on the
“theory of limits”. hese analyses would provide tests to verify some of the
existing theories and frameworks, and to shed some light on the idea of the
“limits of secularization”.
2. Assessing Secularization Trends
As the current state of research does not allow for a more accurate assessment of secularization trends, (except for some local studies; cf. De Graaf and
Te Grotenhuis 2008, Marzano 2010), we propose to explore the most reliable
data provided by European agencies of social survey, in order to shed some light
on the present state of religious decline. he analysis of the data aims both to
ascertain whether a threshold in that process may be identiied, and to establish
typologies or patterns of more or less secularizing trends in diferent societies.
he most reliable source of data regarding the religious state of a large
group of European countries is ofered by the successive waves of the European Values Study (EVS), covering the years of 1981, 1990, 1999-2000, and
the European Social Survey (ESS), covering four periods: 2002, 2004, 2006,
2008. hese surveys present several advantages, as the questionnaires used
are highly standardized, and the questions are the same for the diferent
countries, all of which allows for comparative analysis. As these questionnaires have been collected on a regular base since 1981, the relevant data allows
for a sort of “longitudinal” study.
A sample of 10 European countries was selected among the more than
20 available: France, Italy, Spain, Portugal, Belgium, Britain, Ireland, Germany, Poland, and Netherlands. he sample covers the biggest nations with
mainly Catholic populations, or with a signiicant presence of Catholics.
Britain and Netherlands count with signiicant Catholic populations, even if
they are not the majority religious group. his selection renders the sample
more homogeneous and avoids the pitfalls observed in many cross-national
comparative studies.
he number of case studies collected in each single country by these
surveys is rather limited. he average in the EVS (1999) was 1370 cases for
each of the selected countries, whereas for the ESS (2006), it was 2100.
he religious indicators of these questionnaires are easily identiied. he
EVS ofers a subset on “Religion and Moral”, with 23 items, comprehending “Personal Religious attitudes”, “Religious ailiation and practice”, “Religious atmosphere in family”, and “Moral and Religious Beliefs”. We have
selected three items, with the following headings:
Is there a limit to religious decline?
249
•
“Apart from weddings, funerals and christenings, about how often do
you attend religious services these days?”
•
“How often do you pray to God outside of religious services?”
•
How important is God in your life?” (on a scale of 1 to 10)
he main reason for selecting only these three items is to facilitate comparative research, since the second source of data, the ESS, is less abundant
on religious indicators, providing only these three.
he ESS items on religiosity are located in the subset “Subjective wellbeing, social exclusion, religion, national and ethnic identity”. Apart from
some items related to religious ailiation, the three relevant items for our
research have the following headings:
•
“Apart from special occasions such as weddings and funerals, about how
often do you attend religious services nowadays?”
•
“Apart from when you are at religious services, how often, if at all, do
you pray?
•
“Regardless of whether you belong to a particular religion, how religious would you say you are?” (on a scale of 1 to 10)
As may be appreciated, the irst two are practically identical, and therefore allow us to compare igures from both sources and years. However,
we admit to a certain amount of freedom in our adaptation of the data
regarding the third item, as we have chosen to assimilate the results of two
diferent questions: “How important is God in your life?” (EVS) and “How
religious are you?” (ESS). he data had been organized by extracting the
means of the following values:
•
the attendance of religious services at least once a week (summing the
two irst answers “More than once a week” and “Once a week”);
•
daily prayer;
•
the average answer to the question of: “How important is God/ How
religious are you?”.
Gathering these data has enabled us to compose tables that reveal – at
irst sight – the most basic tendencies. However, a deeper approach requires
a more accurate analysis, revealing patterns of decline or its reverse. herefore, with that end in mind, these initial tables have been the object of an
exponential analysis.
250
Lluis Oviedo - Manuel Canteras
Tables 1,2,3 in the Appendix show the outcomes after gathering the selected data regarding indicators of religiosity for 10 main European Countries. he three selected indicators are highly correlated between them, revealing their mutual implication, as expected.
Before going into the detailed analysis, a couple of observations need to
be made. he case of Italy will be excluded from the successive analysis since
the data of the four ESS waves are missing. With Germany, there is a structural change which accounts for the sharp decline after 2002: the EVS surveys
cover only West Germany, while the ESS waves (from 2002 on) include the
former Eastern Germany, with much higher rates of secularization as a result
of prolonged communist rule.
An initial look at the data as revealed by the three indicators under
review suggests that the secularization tendency has been advancing – at
diferent rates – almost everywhere; nevertheless a closer view reveals interesting traits.
It is important to build a model capable of explaining this decrease and
of measuring its pace, thus allowing for short term previsions. To this end
an exponential model has been elaborated, adjusted for each indicator and
country, with the corresponding coeicient of correlation, which measures
the validity of such adjustments.
he adjusted exponential model is decreasing and tends in an asymptotic way to zero, i.e. the starting point is that the secularization trend would
become complete. In this way those signiicant adjustments would show
that the measured parameter is afected by time, or in other words, tends to
decrease. On the other hand, a non-signiicant adjustment could suggest –
but not prove – that the parameter has reached stability in relation to time.
Indeed, in that stabilized parameter we could test whether “a rest” is still left.
he operated contrasts against the hypothesis of “no-rest” would refute this
hypothesis.
On the other hand the contrasts between descending curves allows for
the comparison of the decrease of various speeds. We applied contrasts of
equality of declines deduced from the adjusted exponential model.
Applying these criteria to the outcomes, for the irst item (attendance of
a religious service) the only countries in which P > 0,05 are Britain and Portugal. In other words, in these two cases it could be assumed that the data are
not adjusted to a model of decline, and could even reach some stability. he
remaining seven countries are adjusted to the exponential model of decline,
and as a consequence, this indicator could still decrease further. [See tables
4,5,6, in the Appendix]
Is there a limit to religious decline?
251
Concerning the second item (daily prayer), only Poland follows the pattern of decline (P = 0,047), and could therefore decrease further; the rest of
the countries examined could be close to stabilizing.
For the third index (how religious are you?), only Portugal and Holland
show some lattening of the downturn tendency; the other seven countries
follow a pattern of decline, even if at a slower pace.
Taking into account these results, the available data do not ofer enough
evidence to conirm the main hypothesis inspiring our research: that religious decline could ind some threshold in most advanced societies. he
downturn continues in most of the European countries, with fewer exceptions, at diferent paces. he suggestion that a kind of “spiritual” tendency
could survive is hard to assess, although the contrast between the prolonged
plunge of attendance indexes and the lattening of those of personal prayer
could ofer some insight in this direction. However, the slow decline of the
third indicator does not allow for quick conclusions.
Looking more carefully at the decline curve, there is not a pattern discriminating higher and lower speeds of decline depending on when countries
started this process: countries starting earlier still observe a strong decline,
like Belgium, Germany and Holland – for the indicator of attendance at least – and a slower one for those starting later, like Portugal and Poland. he
case of Britain is a particular one, with its own characteristics.
However, secularization tendencies in Europe afect more the indicator of religious rites attendance, which decreases at greater pace, than daily
prayer and personal self-assessment. However, this is not the case in all nine
countries: Portugal and – more surprisingly – Britain, seem to have reached
some kind of threshold. More relevant is the result that the second item
addressed, relating to prayer levels, seems to have come to a halt in almost
all the countries considered, (except Poland), while the third – self-reported
religiosity – decreases in most of these countries, but at a much slower pace
(the measure values are also diferent!). he strange thing is that personal
prayer resists the religious crisis better than the self-assessment of personal
religiosity.
he general impression our analysis leaves us with is that igures tend to
stabilize at a diferent rates depending on each society, each indicator and on
several other factors that are hard to account for. Table 7 shows where the igures tend to stabilize for each country and item; there are obvious contrasts
between cases: for daily prayer, France stabilize at about 11%, Portugal at
33%, and Ireland at 46%. Historical and cultural factors seem to ofer more
reliable explanations on these diferences.
252
Lluis Oviedo - Manuel Canteras
he “limits of secularization hypothesis” has still to carry the burden of
proof; that limit could be observed only at the level of private prayer, and
– regarding other indicators – in few cases, depending on speciic circumstances. A still hypothetical threshold would be placed at diferent levels, as
is shown in table 7: higher in countries with national, historical and cultural
characteristics favoring a religious attitude, and lower in others. he igures
justify such a distinction: Poland, Ireland, Portugal and Spain would surely
be in a diferent position to the other countries. It is interesting to note, nevertheless, that in some cases, such as Spain, for example, it is possible for a
country to start in the irst typology – more religiously oriented – and inish
in the second group, inding its own new “niche” and minimal threshold.
All this ofers support to the theory of David Martin, by which diferent
societies experience diversiied processes of religiosity and secularization, depending on distinct cultural and historical patterns (Martin 2005).
3. he “limit of secularization” hypothesis: some data
he second objective of the present research is to look for symptoms
revealing the expected negative impact of secularization in both personal and
social realms. he specter of hypothetic social hardship linked to religious
decline is too broad to be covered by a single study. Our aim is to make a irst
exploration of the available data, focusing thereafter more closely on one case
study: Spain, in which the process of secularization has been more abrupt
than in other countries, whereby its consequences may appear more clearly.
Our method studies the available data in order to identify clues that
can conirm or disclaim such a hypothesis. A irst step consists of verifying
a theory that is deeply related to the issue under exam: the traditional view
on the inverted relationship between levels of income and education, on the
one hand, and religious commitment, on the other. Using the data provided
by the third wave of the ESS (2006), a irst approach identiies items that
could be deemed indicators of the two variables: income and levels of study. We then analyze the correlations between these two items, along with
the aforementioned three extracted items revealing levels of religiosity. he
coeicients are extracted at an individual scale, using the entire population
sampled (around 45,000 cases).
he data ofer a good opportunity to test further the incidence of items
revealing “perceived security”, and national levels of inequality. A second
step turns to oicial European data on levels of fertility, divorce and marriage
stability, in order to assess its possible correlations with the levels of religiosi-
Is there a limit to religious decline?
253
ty already available in the ESS (2006), and in doing so reveal possible “ainities”. he correlations in this case are calculated at national levels, using the
igures of the selected variables for each country, and the means of the items
showing levels of religiosity.
he third step examines the speciic case of Spain, analyzing oicial igures of non-religious marriages and divorce ratings, again with the aim of assessing the possible levels of implication between religious and social trends.
a) Secularization, Wellbeing and Education
One version of the traditional secularization theory states that religious
decline is a consequence of increases in economic wellbeing and higher levels
of education. he evidence resulting from correlating – at individual level –
two indicators from the last wave of ESS (2006): “Feeling about household’s
income nowadays” and “Highest level of education” show – at irst sight
– correlations within the expected pattern: higher levels of education and
income – lower level of religion. Nevertheless, the gathered coeicients are
too low to be taken into account, especially regarding the subjective indicator of income (see table 7). Similar conclusions have been provided before,
especially regarding the impact of education on religion (Halman and Draulans 2006).
Trying to conirm the result regarding the impact of income, we turn to
an analysis on national basis of indicators of wellbeing. he Eurostat Agency
provides reliable data on GDP per capita in “purchasing power standards”
(PPS), “i.e. a common currency that eliminates the diferences in price levels
between countries allowing meaningful volume comparisons of GDP between countries” (oicial explanation). A list of 21 countries for which there
are available data on religious levels (ESS 2006) has been used as the basis
for correlating this economic indicator. he outcomes show no signiicant
correlation in any of the three religious items (see tables 8, 9).
he available data ofers an opportunity to test some of the NorrisInglehart thesis. For example, at individual level ESS 2006 (N = 41230) no
evidence is found about signiicant correlations between items describing
levels of “personal security” (“Feeling of safety when walking alone in the
local area after dark”, “How often do you worry about your home being burgled”, “How often do you worry about becoming a victim of violent crime”,
“Subjective general health”) and indicators of religiosity (R ≤ 0,06). However at the cross-national level a minimally signiicant correlation emerges
between a standard indicator of inequality (Gini coeicient measured by
254
Lluis Oviedo - Manuel Canteras
Eurostat) and of religiosity – attending religious services (N = 16, R= 0,549;
p < 0,028). A similar outcome arises when the data source is more recent
(CIA Factbook of 2008): a minimally signiicant correlation between the
Gini coeicient and the indicator of prayer (N = 21, R = 0,466; p < 0,033);
the correlations are lower and not signiicant when the Gini coeicient is
provided by UNDP (United Nations Development Program) for 2007-8,
and all other indicators of religiosity in Europe.
hese results admit diferent readings, yet a legitimate one is that the
new igures challenge the traditional connection between social progress (in
the form of modernization, rationalization, security and equality) and religious decrease, at least when a certain threshold is reached. It is conceivable
that once certain economic and educative levels are reached, religious commitment does not go down, but stabilizes instead. his is a point already
made with strong evidence by Kaufmann, in contrast with Norris and Inglehart (Kaufmann 2008). In similar terms, it could be stated that at a given
point of social evolution, if religious crisis is no longer considered to be a
consequence of social progress, such social improvement does not imply the
sacriice of religious faith and commitment. his needs to be taken into account if the described theories on the limit of secularization are right. And
on the contrary, such theories would fail if social progress in several ields still
results in the cutting of levels of religious intensity.
b) Family Stability, Fertility, and Religious Involvement
he idea of a relationship between falling religious igures and symptoms
of social and family crisis can be measured comparing a number of variables.
Tables 9 and 10 in Appendix show rates of fertility, marriage, divorce (plus
an added column of marriage without divorce), and the habitual three indicators of religiosity, measured at national levels. he table of correlations
shows at least two signiicant results: there is a clear negative correlation
between the number of divorces and religious attendance, and a positive
one between levels of daily prayer and the diference between marriages and
divorces, (which could be taken as an index of “surviving marriages”).
It must be recognized that reading correlations has a highly hermeneutic dimension: it depends on personal interests, cultural backgrounds and
one’s own research program. However, at a irst sight, a “Weberian” ainity
can be perceived between two of the variables considered: the loss of religious commitment may be associated with problems in family life. Further
analysis involving more variables would be needed in order to better assess
Is there a limit to religious decline?
255
this tendency. It seems obvious that variables of income, educational levels
and other similar factors could load the indicators of familiar instability. A
“pack” of modernizing tendencies seems to be associated with higher rates
of divorces in Europe, a pack in which religious decline constitutes just one
more element among several others.
If these trends can be conirmed, a social scenario could be contemplated in which secularization trends appear not just as an issue concerning
the “religious system”, but also as a limit threatening other social systems. In
that case, the thesis of the “limits of secularization” could be formulated as a
theory of the limits of the entire system and its integration.
In any case, the observed correlations can be understood in both directions, since, irstly, instable families seem to become less religious, and
children of broken families go through a harder time concerning their religious socialization; and secondly, lower religious levels seem to afect family
stability. Indeed, in broad strokes, the data shows a lower religiosity level
among divorced people than amongst those regularly married (divorced
people attend religious services less than half the average of married people,
ESS 2006).
he issue of the possible relationship between religious decline and increasing symptoms of “collective uneasiness” needs to be more deeply investigated. he fertility index seems unafected in the considered model. However, other researches have found signiicant positive correlation between
religious commitment and fertility in trans-national samples and comparing
populations of more and less religious groups (Adsera 2006a,b; Kaufman
2006, 2007, 2008; Hackett 2008, ). Kaufmann has made a good case for
the “absolute limit of secularization”. However fresher data ofers reasons to
doubt his position. A recently published study (Myrskyla, Kohler & Billari
2009) points to a sort of “reverse fertility” process: advanced countries apparently recover from a too low fertility rate. he outcome in our exploration
of the European sample shows a similar pattern: Purchasing Power Standard
(PPS) correlates positively with the fertility rate on a sample of more 25
countries (R = 0,559; p ≤ 0,004). his could be a symptom conirming the
theory of “readjustment” to a non-religious society, and a blow to the limitsof-secularization hypothesis. Nevertheless, the issue is still highly problematic, since many advanced societies are still far from the demographic “replacement level”, despite their “reverse” dynamics. Furthermore, we must take
into consideration the fact that several rich countries strongly subsidise child
bearing with generous beneits, adding a new factor beneath the religious
256
Lluis Oviedo - Manuel Canteras
motivation, and even ofering a “substitute” to it. A related issue is how long
these social policies may be sustainable in the long term.
Other problems could be looked into, such as indicators of youth crises,
and igures of school failure in the ield of education; or igures of mental
health, civic values, and increases of criminality igures. In any case, the possible “ainities” between both forms of decline beg for deeper and broader
assessment.
c) A Closer Look to the Spanish Case
Some national cases merit particular attention in order to better understand secularization dynamics. he Spanish case is highly signiicant in
this respect. Two considerations are worthy of close examination. Firstly,
the igure of non-religious weddings is increasing sharply in the latter years.
his data can be taken as a good indicator of real secularization, more than
self-assessment surveys with limited samples of population. It is therefore
possible to deduce that Spain has not yet reached a minimal threshold in its
secularizing tendency. Secondly, the divorce rate has sharply increased to a
peak of more than 3,39 divorces per 1000 habitants in 2005. Besides this
fact, the correlation between both igures – civil weddings and divorces – is
very high (N = 18; R = 0,892; p < 0,0005); this is surely a relevant outcome
in our context.
his issue deserves deeper analysis. Even if the divorce rate has reached
alarming igures in Spain, it seems, irst, that it goes in line with other more
modernized European societies, or in other words, Spain would reach some
average level in the European context. hen, after 2006 the igures show
some reverse, and appear to adjust more closely to the mean of north and
western European countries (about 2,5 per 1000 inhabitants, after Eurostat
igures). A similar trend can be observed in several other countries of this
area: the rate of divorces tends to stabilize. his outcome could give reason
to both views on the future of religion. In some cases there is a coincidence
between the lattening or deceleration of either, the rate of divorces and the
religious decline. However it could be possible that the advanced societies
tend to adjust to a milieu without religion, and manage to break the decomposing rhythm associated with family instability. he fact that the rate
of civil weddings continues to increase during the last three years could be
an indication that divorce rates are correcting its dangerous increment independently of religious drift. However, some related data awakens further
doubts on the current trends. One example is the steady decline of the total
Is there a limit to religious decline?
257
number of weddings; its rate per 1000 habitants has fallen from 5,61 in
1991, to 4,26 in 2008 (no reverse in this case!); this data can be taken again
as a symptom of disengagement in young generations, discouraged to form
families, and sending negative signals about the future of this society. he
alternative reading could ind its place too: again an episode of transition
and adjustment, in which new models of family and fertility are coniguring
a diferent pattern of social structure. However, it seems in this case that the
last interpretation carries the burden of proof, since it appears less plausible.
[See Table 11.]
Some analysts point to the economic decline experienced in Spain in
the last two years (2007-2009) as a closely related variable afecting the decline of divorces. he igures do not allow for easy conclusions: there is a clear
correlation between increases of rent in the population and other positive
economic indicators until 2008, and the divorce rates. However, the moment of economic deceleration comes much later (2 years) than the timing
of inverse divorce rates.
Other data may be helpful. If we take the number of practiced abortions as a negative social indicator, the Spanish case shows a similar pattern; a steady increase from 1991 (4,83 abortions for 1000 fertile women)
to 2008 (11,54), highly correlating with one of the most reliable indicators
of religious decline – rate of non-religious weddings (N = 18; R= 0,971; p
≤ 0,0005), an outcome contradicting some optimistic assumptions in the
current debate (Paul 2006). However the oicial sources remind us that 43%
of the cases in 2008 were related to immigrant women.
When alternative variables are checked, the outcomes are still less conclusive. he available data does not conirm other expected signs of social or
personal diseases: the fertility rate seems unafected by the variables under
exam; neither do suicide rates appear to be inluenced by religious decline.
Other parameters, such as school failure or rate of psychic pathologies are
very hard to assess, as oicial igures relate often to administrative policies –
in one case – and to the extension of psychic health care – in the other. he
demographic factor represented by the presence of an increasing number of
immigrants, from very diferent cultures and with other agendas and needs
surely distort any prediction or model regarding the religious factor.
he variable represented by what could be termed “troubles with adolescence”, describing increasing diiculties regarding their education, and
more apparent signs of young uneasiness (alcohol and drug consumption,
bullying, violence…) deserve more attention. A more thorough survey
would be helpful, for example taking into account – in a sample of high
258
Lluis Oviedo - Manuel Canteras
schools – the annual rates of disciplinary reports, or denounces of grave bad
behavior among their students. In any case, listening to the opinions of teachers and psychologists involved in these schools, the reasons for such bad
behavior has to do in many cases with the state of the student’s families, and
how de-structured they are. How much such dysfunctions can be related to
religious decline, and how many other variables could be involved, is a question to be better assessed, in Spain and in other countries.
All this needs to be contextualized in relation to other European countries. Available studies show a clear parallelism in Italy regarding the igures of
civil weddings, and increasing divorces and separations: the pattern is the
same as the one observed in Spain (Ferro and Salvini 2007). Whether or
not this trend will stabilize or bottom out, and whether this threshold will
be related to a deceleration in religious decline, is something that needs to
be observed in the next few years, when the registered trends would become
more apparent.
It is too early to say how much of this tendency can be related to the
secularization trend, and how other possible factors are associated to that
crisis. In any case, the igures invite recognition of a critical state in family
stability, linked to more reliable indicators of secularization. he question
arising from this outcome is how much family instability is a society able to
bear before it reacts and goes in search of religious support or an alternative
help so as to correct things that have gone wrong. Furthermore, taking into
account the complexity of current trends, a pressing question is how much a
religious revival could restore better igures of family stability.
4. Provisional Conclusions
he two lines of our study converge in the goal of assessing how far secularization tendencies can go, whether they may reach some kind of “threshold”, and whether such a threshold has to do with the hypothetic negative
consequences associated with religious decline. he theoretical framework
points towards such a limit, as human societies seem to require some minimal levels of religious provision in order to organize a whole series of personal and social needs as well as coping with certain kinds of contingencies
and hazards.
he evidence gathered ofers a complex panorama in which the most
we can expect is some insight into how things currently proceed, and about
some probable links between religious and social crises. Since most of contemporary science has a probabilistic format, such an outcome should be
Is there a limit to religious decline?
259
not dismissed, as it may help, nevertheless, to advance our knowledge of
secularization, and provide orientations for further research.
It is relatively easy to build models trying to describe the hypothetical
trends of religion, which could be represented in a cyclical way throughout
the diferent stages of modernity. A model based on the notion of “relay” is
conceivable: it describes religious dynamics as cycles of growth and decline
– when the decline reaches a minimal threshold, the relay mechanism comes
into play and religious interest increases, in a kind of reactive wave. Sometimes it happens as a consequence of historical or contingent events, but at
other times it could be the result of increasing distress associated with a loss
of meaning and values.
In any case, the suggested model is built at a rather hypothetical level,
where some assumptions are taken for granted, such as: people need religion,
and religion still provides some useful social functions, not yet replaced by
other agencies. A complementary view suggests that religion constitutes a
factor of equilibrium in social dynamics: sometimes a society can do without, but at other times it is required as a way to compensate apparent
unbalance in other ields. In a similar vein, religion can be seen as a kind
of “social bufer”, or an agency aimed at subduing or softening tension and
shocks generated at diferent levels of the social system, or between diferent
sub-systems. When there are no shocks or traumas, the “bufering device” is
not required, but just in case it should be available to perform its task, and
to provide new stability.
Some of the theories exposed at the beginning of our paper may be
brought together to suggest a fairly complex pattern: religion in advanced
societies becomes rather implicit, latent or “vicarious” (Davie 2000) during
times when its presence or function is least required, and can re-emerge in
more explicit forms when its performance becomes more urgent, in response
to a kind of social “relay” or as a bufer mechanism aimed at deterring the
most disruptive tendencies.
Perhaps the Rational Choice theory applied to the study of religion can
contribute to this theoretical framework. In a similar way to economic analysis, religious demand will increase when greater scarcity of values claims the
provision of social certainty, and decrease when such values are deemed unnecessary, or when the costs of religious practice do not compensate for the
expected rewards. After all, the mainstream religions are about “salvation”:
when salvation is less required, due to the fact that societies and individuals
manage to arrange their life in an acceptable and satisfactory way, religious
supplies seem less essential. he same applies the other way around: when
260
Lluis Oviedo - Manuel Canteras
secular means do not manage to satisfy personal and social needs, religious
provision of meaning and certainty would cover that demand (Sosis 2007).
he current situation of economic crisis could provide a good case study to
test the validity and heuristic power of this theory, when enough time and
data could be gathered.
In principle, the idea of a limit in the secularization process does not necessarily contradict the standard theories explaining this tendency; however
it adds an element that implies a revision of the existing trends and of projection models. Furthermore the available data appears as compatible with the
idea of “dormant religion”, awakening only in moments of special hardship.
In the opening pages we pointed to the existence of at least a powerful
theory contradicting the theses on a threshold in the process of secularization. David Voas suggested this view in a personal exchange: we are just
witnessing the consequences of a moment of adjustment not completed and
to its negative efects; however advanced societies going through this process
could ind a way to adapt to the new circumstances, as happens in every evolutionary process. Even if we consider the scarce consequence of contextual
factors of wellbeing in religious decline, and the negative efects of decreasing religion on family formation and stability, the analysis of the available
data, at the current state of research, does not supply complete evidence to
support the main hypothesis on the limit of secularization. he scientiic study of religion seems unable to decide convincingly on any one of the theories
disputing the issue of religion and social uneasiness. he recent igures are
susceptible to alternative readings, since most of the gathered data is inconclusive and still open to uncertain developments.
Part of the problem is related to the limits of the available data. Some
colleagues have lagged up deiciencies in the EVS and ESS, as the samples
are too reduced and the data was often gathered by diferent companies in
diferent ways during the various survey years, especially in the case of the
EVS. All this works against an excessively decided use of the data to build
longitudinal studies. he accuracy and reliability of these surveys is rather
limited, but at the moment, it is all we have available to provide an international basis for comparison.
A better case could be built by resorting to other sources of national
data, as in the Spanish case, where more objective indicators of religiosity can
be obtained, that go beyond subjective self-assessment. A good research program would have to aim towards overcoming these revealed shortcomings,
in order to present better diagnoses of the religious situation and its evolution through the years.
Is there a limit to religious decline?
261
Critics could object furthermore to a method using in an undiferentiated way individual approaches, national, and cross-national to build our own
case. We tried to explore the diferent scenarios looking for every symptom
that could conirm our working hypothesis. Obviously these diferent approaches – rather explorative – need to be reined in future elaboration.
In any case these problems do not afect just our exploration, but project
a shadow on every projection of future tendencies of the European religious
scene, which appear as very hypothetical. Some outcomes are to be expected,
based on similar data to those applied in our research (Voas 2009), but since
religious behavior is so context-bound and sensitive to the surrounding environment, its evolution seems to be rather contingent and hard to predict.
Some inal considerations emerge rather as a relexive outcome after
months of dealing with the raw data. he irst one concerns the unavoidable
amounts of contingence and uncertainty involving the study of religion, its
present and future trends; to be aware of this means to renounce to any security about future trends, both for secularists or for those foreseeing some
kind of preservation of religion; in our view the horizon remains very open,
and it is hard to state any secure tendency, or even any kind of threshold. he
second relection is aimed at the problematic relationship between theory
and the empirical ield in social and human sciences. Despite the amount of
theories that could be understood as supporting the idea of a limit on religious decline, the reality – so far – does not conirm in certain terms these
theoretical corpus; at most we can correct views which point too quickly to
the religious inevitable decline, and suggest “some probability” to the link
between religious fall and personal or social diiculties; no more, no less. he
last relection is addressed to theologians and philosophers of religion: we
need a more “empirical” or “experimental” theology in order to make sense
of these outcomes, and to overcome the aprioristic ways currently presiding
over these disciplines. It is hard to state any longer a priori the “universal
need of religion”, as an axiom many theorists of religion were relying on. As
a consequence, nobody should be too conident expecting a “return of religion” or a “rebound” after touching the threshold of “minimal religiosity”.
Religious experience seems to depend more in advanced societies – for good
or ill – on more free factors, not constrained by absolute limits or boundaries, and so, its ofering and practice could be better conceived on this openended ield of human freedom, by its reasons and feelings.
Closing our analysis, let us remind ourselves that theology must assume
the decisive role of “whistle blower” when things are going really bad in our
social and ecclesial context. heologians, however, are not called to just ring
262
Lluis Oviedo - Manuel Canteras
the alarm bells to alert everyone else to the incoming risks and dangers. Since
its main function is to relect on the relevance of faith, taking into account
the threats to the credibility of the Christian message is the irst step into a
process leading to the proposal of better arguments and appropriate answers.
Data References
European Values Study (EVS) 1st wave 1981 [Integrated data ile ZA4438] 2006
/ Release 2, Spain, Germany: ASEP/JDS, GESIS Data Archive and Data Analysis,
Cologne, [producer]. Germany: GESIS Data Archive and Data Analysis, Cologne
[distributor], Downloaded [12.10. 2009] from ZACAT
http://zacat.gesis.org/webview/index.jsp?object=http://134.95.45.58:80/obj/fStudy/ZA4438
European Values Study 2nd wave 1990 [Integrated data ile ZA4460] 2006 /
Release 2, Spain, Germany: ASEP/JDS, GESIS Data Archive and Data Analysis,
Cologne, [producer]. Germany: GESIS Data Archive and Data Analysis, Cologne
[distributor], Downloaded [12.10.2009] from ZACAT
<http://zacat.gesis.org/webview/index.jsp?object=http://134.95.45.58:80/obj/fStudy/ZA4460>
European Values Study 3rd wave 1999/2000 [Integrated data ile ZA3811] 2006 /
Release 2, he Netherlands, Germany: Tilburg University, GESIS Data Archive and
Data Analysis, Cologne, Netherlands Institute for Scientiic Information Services
(NIWI), Amsterdam [producer] 2003. Germany: GESIS Data Archive and Data
Analysis, Cologne [distributor]. Downloaded [12.10.2009] from ZACAT <http://
zacat.gesis.org/webview/index.jsp?object=http://134.95.45.58:80/obj/fStudy/
ZA3811>
European Social Survey (ESS) round 1
R Jowell and the Central Co-ordinating Team, European Social Survey 2002/2003;
ed. 6.1: Technical Report, London: Centre for Comparative Social Surveys, City
University (2003). Norwegian Social Science Data Services (NSD) as the data archive and distributor of the ESS data. Downloaded [12.10.2009] from Nesstar
http://nesstar.ess.nsd.uib.no/webview/?v=2&mode=documentation&submode=abs
tract&study=http://129.177.90.164:80/obj/fStudy/ESS1e06.1&top=yes
ESS round 2
R Jowell and the Central Co-ordinating Team, European Social Survey 2004/2005;
ed. 3.1: Technical Report, London: Centre for Comparative Social Surveys, City
University (2005). Norwegian Social Science Data Services (NSD) as the data archive and distributor of the ESS data. Downloaded [12.10.2009] from Nesstar
Is there a limit to religious decline?
263
http://nesstar.ess.nsd.uib.no/webview/?v=2&mode=documentation&submode=abs
tract&study=http://129.177.90.164:80/obj/fStudy/ESS1e06.1&top=yes
ESS round 3
R Jowell and the Central Co-ordinating Team, European Social Survey 2006/2007;
ed. 3.2: Technical Report, London: Centre for Comparative Social Surveys, City
University (2007). Norwegian Social Science Data Services (NSD) as the data archive and distributor of the ESS data. Downloaded [12.10.2009] from Nesstar
http://nesstar.ess.nsd.uib.no/webview/?v=2&mode=documentation&submode=abs
tract&study=http://129.177.90.164:80/obj/fStudy/ESS1e06.1&top=yes
ESS round 4
R Jowell and the Central Co-ordinating Team, European Social Survey 2008/2009;
ed. 3.2: Technical Report, London: Centre for Comparative Social Surveys, City
University (2007). Norwegian Social Science Data Services (NSD) as the data archive and
distributor of the ESS data. Downloaded [12.10.2009] from Nesstar
References
Adsera, Alicia. 2006a. “Marital Fertility and Religion: Recent Changes in Spain.”
Population Studies. 60 (2): 205-221.
Adsera, Alicia. 2006b. “Religion and Changes in Family-Size Norms in Developed
Countries.” Review of Religious Research. 47 (3): 271-286.
Atran, Scott. 2002. In Gods We Trust: he Evolutionaly Landscape of Religion. Oxford,
New York: Oxford University Press.
Berger, Peter. L., Brigitte Berger and Hansfried Kellner. 1974. he Homeless Mind:
Modernization and Consciousness. London: Penguin.
Berger, Peter L. 1999. he Desecularization of the World: A Global Overview. In
he Desecularization of the World: Resurgent Religion and World Politics, ed.
GWe al. pp. 1-18. Grand Rapids, MI: Eerdmans.
Bergmann, Sigurd (2003). God in Context: A Survey of Contextual heology, Aldershot, UK, Burlington, VT: Ashgate.
Bevans Stephen B. (1992). Models of Contextual heology, Maryknoll - New York:
Orbis.
Bruce, Steven. 2002. God is Dead: Secularization in the West. Oxford - Malden, MA:
Blackwell.
Casanova, José. 1994. Public Religions in the Modern World. Chicago: University of
Chicago Press.
Davie, Grace. 2000. Religion in Modern Europe: A Memory Mutates. Oxford - New
York: Oxford Univ. Pr.
264
Lluis Oviedo - Manuel Canteras
De Graaf, Dirk Nan and Manfred Te Grotenhuis. 2008. “Traditional Christian
Belief and Belief in the Supernatural: Diverging Trends in the Netherlands
Between 1979 and 2005?” Journal for the Scientiic Study of Religion. 47(4):
585-598.
Dobbelaere, Karel. 2002. Secularization: An Analysis at hree Levels, Bruxelles, Bern
and Berlin: Peter Lang.
Ferro, Irene and Silvana Salvini. 2007. “Separazione e divorzio in Italia. Le tendenze
e le diferenze regionali.” Popolazione e Storia. 2007 (1): 125-155.
Frejka, Tomas, and Westof, Charles F. 2006. “Religion, Religiousness and Fertility
in the U.S. and in Europe.” Working Paper WP-2006-013. Rostock, Germany:
Max Planck Institute for Demographic Research.
Giddens, Anthony. 1991. Modernity and Self-Identity: Self and Society in the Late
Modern Age, Cambridge: Polity Press.
Gill, Anthony and Erik Lundsgaarde. 2004. “State Welfare Spending and Religiosity.” Rationality and Society. 16 (4): 399-436.
Goldstein, Warren S. 2009. “Secularization Patterns in the Old Paradigm.” Sociology of Religion. 70(2):157-178.
Greeley, Andrew. 2002. Religion in Europe at the End of the Second Millennium: A
Sociological Proile. New Brunswick, NJ: Transaction.
Hackett, Conrad Peter. 2008. Religion and Fertility in the United States: he Inluence
of Ailiation, Region, and Congregation. Doctor hesis at Princeton University,
2008.
Halman, Loek, and Veerle Draulans. 2006. “How Secular Is Europe?” British Journal of Sociology. 57(2): 263-288.
Halman, Loek, and Veerle Draulans. 2006. “How Secular Is Europe?” British Journal of Sociology. 57(2): 263-288.
Jenkins, Philip. 2007. God’s Continent: Christianity, Islam, and Europe’s Religious Crisis. Oxford - New York: Oxford Univ. Pr.
Kaufmann, Eric. 2006. “Breeding for God.” Prospect Magazine, November, 128.
Kaufmann, Eric. 2007. “he End of Secularisation in Europe?: A Demographic
Perspective.” Working Paper. Available at
http://www.sneps.net/RD/religdem.html, opened 26.09.09.
Kaufmann, Eric. 2008. “Human Development and the Demography of Secularization in Global Perspective.” Interdisciplinary Journal of Research on Religion. 4,
http://www.religjournal.com/pdf/ijrr04001.pdf, opened 26.09.09.
Knobloch, Stefan. 2006. Mehr Religion als gedacht! Wie die Rede von Säkulariesierung
in die Irre führt. Freiburg, Bassel and Wien: Herder.
Luhmann, Niklas. 1989. “Die Ausdiferenzierung von Religion”, in Gesellschaftstruktur
und Semantik: Studien zur Wissenssoziologie der modernen Gesellschaft 3, Frankfurt
a.M.: Suhrkamp.
Is there a limit to religious decline?
265
March, James B. 1994. A Primer on Decision Making: How Decisions Happen. New
York: he Free Press.
Martin, David. 2005. On Secularization: Towards a Revised General heory. Aldershot - Burlington VT: Ashgate.
Marzano, Marco (2010) “he ‘Sectarian’ Church. Catholicism in Italy since John
Paul II”, unpublished paper.
McCauley, Robert N. and E. homas Lawson. 2002. Bringing Ritual to Mind: Psychological Foundations of Cultural Forms. Cambridge and New York: Cambridge
University Press.
Mikko Myrskyla, Hans-Peter Kohler and Francesco C. Billari. 2009. “Advances in
Development Reverse Fertility Declines.” Nature, 460/6: 741-743.
Moreno-Riaño, Gerson, Mark C. Smith and homas Mach. 2006. “Religiosity, Secularism, and Social Health: A Research Note.” Journal of Religion and Society.
8, Available at: http://moses.creighton.edu/JRS/2006/2006-1.html, opened
26.09.09.
Norris, Pipa and Ronald Inglehart. 2004. Sacred and Secular: Religion and Politics
Worldwide. Cambridge - New York: Cambridge University Press.
Paul, Gregory S. 2005. “Cross-National Correlations of Quantiiable Societal
Health with Popular Religiosity and Secularism in the Prosperous Democracies: A First Look.” Journal of Religion and Society, 7. http://moses.creighton.
edu/JRS/2005/2005-11.html, opened 26.09.09.
Pears, Angie. 2009. Doing Contextual heology, London, New York: Routledge ,
Rees, homas J. 2009. “Is Personal Insecurity a Cause of Cross-National Diferences
in the Intensity of Religious Belief?: An Interchange between Psychology and
Religion.” Journal of Religion and Society,11, 2009, at:
http://moses.creighton.edu/JRS/2009/2009-17.html, opened 26.09.09.
Rémond, René. 1999. La secolarizzazione. Religione e società nell’Europa contemporanea. Roma and Bari: Laterza.
Schreiter, Robert (1985). Constructing Local heology, Maryknoll - New York: Orbis
Books.
Sosis, Richard. 2007. “Psalms for Safety: Magico-Religious Responses to hreats of
Terror.” Current Anthropology. 48(6): 903-911.
Stark, Rodney. 1999. “Secularization RIP.” Sociology of Religion. 60: 249-273.
Stark, Rodney and Roger Finke. 2000. Acts of Faith: Explaining the Human Side of
Religion. Berkeley CA: University of California Press.
Taylor, Charles. 2007. A Secular Age. Harvard and London: Harvard University
Press.
Turina, Isacco. (2007). “Secularization as a Property of Action”, Social Compass, 54
(29, 161-173.
266
Lluis Oviedo - Manuel Canteras
Voas, David. 2009. “he Rise and Fall of Fuzzy Fidelity in Europe.” European Sociological Review. 25(2): 155-168.
Voas, David and Alasdair Crockett. 2005. “Religion in Britain: Neither Believing
nor Belonging.” Sociology 39(1): 11-28.
Young, Cristobal. 2009. “Model Uncertainty in Sociological Research: An Application to Religion and Economic Growth.” American Sociological Review. 74 (3):
380-397.
267
Is there a limit to religious decline?
APPENDIX
TABLE 1
MEANS OF ATTENDANCE TO RELIGIOUS SERVICES AT LEAST WEEKLY,
IN 9 MAIN EUROPEAN COUNTRIES
Year
Fr
It
Sp
1981 EVS
10,7
10,2
32,4
37,9
40,1
30,2
1990 EVS
PT
Be
UK
Ir
Ge*
39,1
30,0
26,5
13,4
14,2
82,2
80,9
18,5
17,9
25,4
1997 EVS
Pl
Hl
65,6
25,0
20,2
13,5
56,1
2000 EVS
7,6
40,5
25,5
37,4
18,7
14,5
65,4
15,9
59,2
13,9
2002 ESS
7,7
31,1
21,2
29,1
11,0
13,1
53,9
8,1
55,6
12,3
2004 ESS
6,9
19,0
29,9
9,4
14,0
58,2
8,7
56,7
13,2
2006 ESS
6,9
6,1
18,3
16,2
28,7
32,3
9,2
8,7
12,9
12,2
47,0
8,6
7,9
53,0
52,4
13,1
2008 ESS
* EVS only West Germany; ESS, all Germany
TABLE 2
MEANS OF DAILY PRAYER, IN 9 MAIN EUROPEAN COUNTRIES
Year*
Fr
It
Sp
Pt
Be
UK
Ir
Ge
Pl
Hl
1990 EVS
2000 EVS
12,5
9,4
33,7
37,9
25,5
22,1
26,3
39,2
18,3
19,8
22,9
15,6
48,9
50,5
21,2
10,0
54,2
53,0
24,2
21,4
2002 ESS
12,6
31,4
21,7
35,9
14,4
20,5
48,3
12,6
44,9
26,0
2004 ESS
11,5
18,9
30,0
14,8
19,6
52,6
12,5
44,4
25,6
2006 ESS
10,3
9,7
24,5
19,6
32,3
36,7
15,7
14,8
19,7
19,1
44,1
12,6
13,9
47,6
42,5
23,2
2008 ESS
* For EVS 1990, the answer taken into account to the question “How often do you pray to God outside
of religious services?” was: “often” instead of “every day”.
TABLE 3
AVERAGE OF “HOW IMPORTANT IS GOD/ HOW RELIGIOUS ARE YOU?” IN 9
MAIN EUROPEAN COUNTRIES
Year*
Fr
It
Sp
Pt
Be
UK
Ir
Ge
Pl
Hl
1981 EVS
1990 EVS
4,53
4,44
6,77
7,05
6,32
6,12
7,10
5,84
5,31
5,43
5,37
7,98
7,93
5,33
5,30
8,43
5,17
4,89
2000 EVS
4,37
7,43
5,97
7,83
5,39
4,91
7,74
4,39
8,37
4,92
2002 ESS
3,83
6,08
4,46
5,61
4,96
4,39
5,80
3,87
6,53
5,12
2004 ESS
3,78
4,43
5,34
4,78
4,45
6,0
3,89
6,53
5,04
2006 ESS
2008 ESS
3,70
3,61
4,58
4,51
5,79
5,95
4,92
4,77
4,08
4,05
5,41
3,86
4,02
6,48
6,41
4,89
* EVS question: “How important is God in your life?”, rating from 1-10; ESS question: “How religious
are you?”, (rating from 0-10)
268
Lluis Oviedo - Manuel Canteras
TABLE 4
EXPONENTIAL MODEL OF THE MEANS OF “ATTENDANCE OF RELIGIOUS
SERVICES AT LEAST WEEKLY”
Country
France
Spain
Portugal
Belgium
Britaim
Ireland
Germany
Poland
Holland
Model
Y = exp (44,742 – 0,021X)
Y = exp (67,335 – 0,032X)
Y = exp (34,986 – 0,016X)
Y = exp (104,675 – 0,051X)
Y = exp (7,495 – 0,002X)
Y = exp (46,195 – 0,021X)
Y = exp (73,56 – 0,036X)
Y = exp (25,717 – 0,011X)
Y = exp (60,842 -0,029X)
Correlation
R = 0,97
R = 0,98
R = 0,754
R = 0,937
R = 0,393
R = 0,910
R = 0,864
R = 0,888
R = 0,975
Signiic.
P<0,0005
P<0,0005
P = 0,083
P = 0,002
P = 0,384
P = 0,012
P = 0,006
P = 0,008
P = 0,001
TABLE 5
EXPONENTIAL MODEL OF THE MEANS OF “DAILY PRAYER”
Country
France
Spain
Portugal
Belgium
Britaim
Ireland
Germany
Poland
Holland
Model
Y = exp (23,673 – 0,011X)
Y = exp (26,262 – 0,012X)
Y = exp (-23,537 + 0,014X)
Y = exp (29,51 – 0,013X)
Y = exp (18,341 – 0,008X)
Y = exp (8,381 – 0,002X)
Y = exp (52,623 – 0,025X)
Y = exp (29,394 – 0,013X)
Y = exp (1,502 + 0,001X)
Correlation
R = 0,532
R = 0,624
R = 0,583
R = 0,648
R = 0,391
R = 0,214
R = 0,640
R = 0,668
R = 0,066
Signiic.
P = 0,278
P = 0,185
P = 0,224
P = 0,164
P = 0,444
P = 0,730
P = 0,171
P = 0,047
P = 0,916
TABLE 6
EXPONENTIAL MODEL OF AVERAGES FOR “HOW IMPORTANT IS GOD/
HOW RELIGIOUS ARE YOU”
Country
France
Spain
Portugal
Belgium
Britaim
Ireland
Germany
Poland
Holland
Model
Y = exp (19,153 – 0,009X)
Y = exp (31,370 -0,015X)
Y = exp (30,707 – 0,014X)
Y = exp (15,884 - 0,007X)
Y = exp (32,242 – 0,015X)
Y = exp (32,792 – 0,015X)
Y = exp (29,274 – 0,014X)
Y = exp (37,768 – 0,018X)
Y = exp (3,492 – 0,001X)
Correlation
R = 0,805
R = 0,864
R = 0,632
R = 0,927
R = 0,766
R = 0,862
R = 0,930
R = 0,863
R = 0,371
Signiic.
P = 0,006
P = 0,012
P = 0,178
P = 0,003
P = 0,003
P = 0,027
P = 0,002
P = 0,027
P = 0,469
269
Is there a limit to religious decline?
TABLE 7
RESTS AND INTERVALS OF CONFIDENCE (95%)
OF COUNTRIES FOR EACH ITEM
Country
Attend
Pray
10,93%
12,49%)
21,92%
24,82%)
33,10%
38,65%)
16,18%
18,57%)
19,44%
22,18%)
46,49%
52,93%)
13,42%
17,43%)
France
Spain
Portugal
32,45%
(28,27%; 37,38%)
Belgium
Britain
13,45%
(12,71%; 14,23%)
Ireland
Germany
How Rel.
(9,56%;
(19,36%;
(28,35%;
6,2 (5,3; 7,2)
(14,10%;
(17,94%;
(44,98%;
(10,34%;
Poland
24,02%
48,8%)
Holland
(21,78%;
5,0 (4,9; 5,1)
TABLE 8
PEARSON CORRELATIONS ON ALL THE CASES OF ESS 2006 ROUND AND
RELEVANT ITEMS N= 41341-42570
Items
How religious are you
How often attend religious services
apart from special occasions1
How often pray apart from at religious
services1
Highest level of education2
Feeling about household's income
nowadays3
Relig
1
Attend
-,606**
Pray
-,675**
Educ.
-,122**
Inc.
,069**
-,606**
1
,642**
,113**
-,051**
-,675**
,642**
1
,137**
-,103**
-,122**
,113**
,137**
1
-,193**
,069**
-,051**
-,103**
-,193**
1
** Correlation signiicant at 0,0005 (bivariate),
1. Seven levels inverted: 1= more frequency – “every day”; 7= less – “never”
2. Seven levels, from less to more: 0 = “Not completed primary education”; 6= “Second stage of
tertiary”.
3. Four levels, from more to less: 1 = "Living comfortably on present income”; 2 = “Coping on present income”; 3 = “Diicult on present income”; 4 = “Very diicult on present income”
270
Lluis Oviedo - Manuel Canteras
TABLE 9
FERTILITY, MARRIAGE AND DIVORCE RATES IN EUROPEAN COUNTRIES, PLUS
INDICATORS OF RELIGIOSITY AND PURCHASING POWER 2006
Austria
Belgium
Bulgaria
Croatia
Cyprus
Czech
Denm.
Estonia
Finland
France
German
Greece
Hungary
Iceland
Ireland
Italy
Netherl.
Norway
Poland
Portugal
Roman
Slovakia
Slovenia
Spain
Sweden
Switzer.
U. K.
Fert1
1,40
1,37
1,38
1,47
1,33
1,83
1,55
1,84
2,00
1,32
1,39
1,34
1,46
1,90
1,32*
1,70
1,42
1,27
1,35
1,31
1,24
1,31
1,38
1,85
1,90 Fine
modulo
1,84
Mar2
4,46
4,25
4,26
4,97
6,80
5,15
6,71
5,18
5,36
4,34
4,54
5,18
4,42
8,91
5,13
4,13
4,35
4,31
5,93
4,52
6,79
4,81
3,17
4,61
5,02
Div2
2,5
2,8
1,9
1,1
2,3
3,1
2,6
2,8
2,5*
2,2
2,3
1,3
2,5
1,3
0,8*
0,8*
1,9
2,3
1,9
2,3
1,5
2,4
1,2
1,7*
2,2
4,66
5,23*
M-D3
1,96
1,45
2,36
3,87
4,50
2,05
4,11
2,38
2,86
2,14
2,24
3,88
1,92
7,61
4,33
3,33
2,45
2,01
4,03
2,22
5,29
2,41
1,97
2,91
2,82
Att4
16,0
9,2
5,8
Pray4
18,0
15,7
8,0
HowRel4
16,0
39,1
5,10
3,6
2,5
5,4
6,9
8,6
10,9
7,6
21,1
10,3
12,6
4,29
3,58
5,29
3,70
3,86
11,9
21,7
4,41
47,0
44,1
5,41
13,1
5,5
52,0
28,7
23,2
13,5
47,6
32,3
4,89
3,81
6,48
5,79
31,1
18,1
18,3
4,2
30,2
16,2
24,5
10,1
5,90
4,69
4,58
3,55
PPS5
124,3
118,5
36,5
51,7
90,3
77,4
122,9
65,3
114,9
109,5
115,8
94,1
63,6
123,7
147,4
103,5
130,9
183,7
52,3
76,4
38,4
63,5
87,7
104,1
121,5
2,3
2,36 11,6
30,4
5,50
136
2,6*
2,63 12,9
19,7
4,08
120,4
Source of fertility, marriages, divorce and PPS data: Eurostat 2006
* Data for 2005
1. Number of children per woman
2. Marriages and divorces per 1000 persons
3. Number of marriages minus divorces
4. Data from ESS 2006
5. GDP per capita in Purchasing Power Standards (PPS)
6. Gini Coeicient of inequality 2005 Eurostat
5,10
4,92
4,30
Gini6
26
28
29
26
24
34
26
28
28
33
28
32
33
27
36
41
31
26
32
23
271
Is there a limit to religious decline?
TABLE 10: PEARSON CORRELATIONS AMONG THE ABOVE MENTIONED
FACTORS ON COUNTRY BASE; N= 20-27
Fert.
Fertil.
Marria
Marri.
Div.
M-D
Att.
Pray
H. Rel
PPS
Gini
1
,062
,156
-,020
-,273
-,132
-,260
,546**
-,360
,062
1
-,057
,881**
,129
,351
,170
-,055
-,049
-,297
Divor
,156
-,057
1
-,52**
-,55**
-,399
-,226
,052
M-D
-,020
,881**
-,522**
1
,445*
,573**
,296
-,071
,196
Atten
-,273
,129
-,554**
,445*
1
,864**
,767**
-,211
,549*
,502*
Pray
-,132
,351
-,399
,573**
,864**
1
,835**
-,096
How. Rel
-,260
,170
-,226
,296
,767**
,835**
1
-,239
,410
,546**
-,055
,052
-,071
-,211
-,096
-,239
1
-,397
-,360
-,049
-,297
,196
,549*
,502*
,410
-,397
1
PPS
Gini 1
** Correlation signiicant at 0,01. * Correlation signiicant at 0,05.
TABLE 11: RATES OF WEDDINGS, DIVORCES, FERTILITY, SUICIDES AND
ABORTS IN SPAIN
Year
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
Rate NonRel Wedding1
20,89
20,63
22,36
22,7
22,48
23,07
23,39
24,07
24,18
24,14
26,64
28,91
33,06
37,03
39,01
42,71
45,17
48,74
% Di- Ratio Wed- Ratio Div
Fertility
vorce1 ding Total2
Total2
Rate3
30,74
5,61
1,73
1,33
30,7
5,56
1,71
1,32
35,95
5,06
1,82
1,27
39,63
4,96
1,97
1,2
41,15
4,96
2,04
1,17
43,28
4,89
2,12
1,16
45,29
4,95
2,24
1,17
45,4
5,2
2,36
1,16
46,38
5,18
2,4
1,19
47,35
5,34
2,53
1,23
49,63
5,06
2,51
1,24
54,5
5,06
2,76
1,26
59,79
4,97
2,97
1,31
62,51
5
3,13
1,33
71,76
4,72
3,39
1,35
71,72
4,55
3,26
1,38
68,21
4,46
3,04 .
60,49
4,26
2,58 .
Source: Instituto Español de Estadística (INE)
Rates per hundred celebrated weddings in the same year
2
Rates per 1000 habitants
3
Rates per 1000 fertile women
4
Rate per 10000 people
1
Abort
Rate3
4,83
5,14
5,16
5,39
5,53
5,69
5,52
6
6,52
7,14
7,66
8,46
8,77
8,94
9,6
10,62
11,49 .
11,54 .
Suicide
Rate4
6,66
6,06
6,17
6,98
6,91
6,91
6,98
6,6
6,23
6,35
5,5
5,5
5,26
5,32
5,04
4,03