UN Population Division, Department of Economic and Social Affairs,
with support from the UN Population Fund (UNFPA)
|
| UN Population Division, Department of Economic and Social Affairs, with support from the UN Population Fund (UNFPA) |
|
***********************************************************************
This document is being made available by the Population Information
Network (POPIN) Gopher of the United Nations Population Division,
Department for Economic and Social Information and Policy Analysis, in
collaboration with the European Association for Population Studies and
the IUSSP. For further information please contact Professor G.C.
Blangiardo, Local Organizer, EAPS Conference, Milan, University of Milan,
Istituto di Statistica, V. Visconti di Modrone 21, Milan, Italy.
************************************************************************
EUROPEAN POPULATION CONFERENCE
CONGRES EUROPEEN DE DEMOGRAPHE
Milano, 4-8 settembre 1995
Plenary Session VII
Is there a new conservatism that will bring back the old family?
Ideational trends and the stages of family formation in
Germany, France, Belgium and the Netherlands, 1981-1990
by Ron Lesthaeghe and Guy Moors
1. Introduction
The unfolding of the life course in Western societies with
respect to home leaving, union formation, parenthood and further
family building has been connected to a variety of factors, both of
an economic and ideational nature. A rough classification of these
factors could be as follows:
a. Factors dealing with human capital and with the
opportunity structure
These factors concern economic effects that are either
associated with female education and employment, or with the career
possibilities of younger cohorts and with relative deprivation
evaluated against consumption aspirations developed during the
socialization phase [1]. According to the human capital theory,
increased female economic autonomy reduces the female dependence on
(re)marriage and enhances the opportunity costs associated with
parenthood. The economic relative deprivation theory posits that
the postponement of all major transitions concerning family
formation stem from the frustration of consumption aspirations
among incoming cohorts, which have been facing, especially since
the 1970s, more difficulties in career development. The neo-
classic human capital theory predicts no change in the pattern of
late family formation for as long as female human capital remains
valued. The relative deprivation theory, on the other hand, leaves
the possibility for a return to earlier patterns of family
formation if opportunity structures for incoming smaller cohorts
would improve.
Not only the economics of the family deal with such issues.
Also sociological perspectives concerning the marriage market
connect the advent of new forms of partner selection (e.g. via
premarital cohabitation) to rising female education and
concommittant changes in tastes [2]. b. Factors associated with
selection effects stemming from the family of origin and the
socialization phase
A number of studies deal with the economic or educational
background of the parental household [3] and another group of
studies pays attention to the psychological climate in the family
of origin, the experience with parental divorce, a single parent or
with reconstituded families [4]. The first set of studies mainly
focuses on the selection effects of social class and the innovation
of new forms of demographic behaviour. For instance, in which
contexts was there an emergence of premarital cohabitation among
the sons and daughters of the ®bourgeoisie¯, and in which settings
was there an early spread among the ®working class¯? [5] The
second set of studies, dealing with the psychological conditions in
the family of origin, point out that earlier home leaving and
premarital cohabitation are considerably more likely to occur among
young individuals who have experienced a parental divorce or a
reconstituted family.
Another set of selection effects are ideational in nature.
Choices with respect to both family formation and career
orientation are seen as stemming from the values developed during
the ®formative years¯ [6]. In this respect, the recurrence of a
strong selection effect stemming from a religious upbringing has
been repeatedly documented, along with the selection effects of
other ethical, political and role orientations developed earlier in
life [7]. In this view value orientations are not merely
correlates of the economic aspects of career developments, and
hence merely endogenous or even tautological byproducts of an
economic theory, but basic ingredients of utility functions and
life-style orientations which guide individuals and cohorts through
much of their adult life. In other words, the ideational theory
expects the existence of major cohort effects, and assumes that
societal change takes place according to N. Ryder's model of
®social metabolism¯ [8]. c. Factors connected with the dynamics of
the life-course unfolding itself
The factors are embedded in models that posit the existence
of major recursive effects between value orientations, cost-
benefit calculation and actual decisions. In such dynamic models,
psychological mechanisms are operative so that values are adapted
to life-course unfolding. Such adaptations can occur a priori or
a posteriori. The former type emerges when values are altered in
anticipation of a decision which is likely to be made in the near
future. Such mechanisms are described by the notions of
®anticipatory socialization¯ and ®reference group behaviour¯. When
values are adjusted to prior decisions, we are obviously dealing
with ex post rationalizations, or with affirmation or negation
effects. In the instance of affirmation, values which were held a
priori and caused selection toward a particular decision are
reinforced by that decision. In the case of negation, the opposite
occurs and the individual reverses his value ranking.
The vanguard of current research undoubtedly lies in the use
of such recursive models, but their quantification requires
detailed panel studies [9]. The selection effect of value
orientations can then be measured by relating events occurring in
the window between two waves to the value measurements performed in
the first wave. Conversely, events in the window can be used to
predict the presence or lack of value changes as measured in the
two successive waves. Changes in values during the life course are
often associated with the arrival of a third person such as a
partner or a child. In these instances value changes result from
the ®give and take¯ processes between partners, from bargaining
leading to joint decisions, and from the internalization of the
well being of others in one own's utility function. Connections
with social exchange theory and the notion of altruism become
evident [10]. Moreover feed- back mechanisms accross generations
are possible with the older cohorts adapting their views and
behaviour (e.g. post-marital cohabitation) to what the younger
cohorts are articulating and implementing. Such cases of ®reverse
socialization¯ are equally documented [11].
2. The scope of the present paper
This paper only addresses two issues:
1. the statistical association between the religious, ethical
and political orientations of individuals on the one hand and life
course events such as home leaving, cohabitation, parenthood and
progression to higher parities on the other hand;
2. the trends in the ideational correlates themselves between
1981 and 1990 in each of the cohorts.
The issue of associations between ideational and
demographic factors has already be documented elsewhere for the
European Values Survey (EVS) samples of Germany (West), France,
Belgium and the Netherlands [12]. But this subject needs to be
revisited here for values measurements that are comparable in the
two EVS-samples of 1981 and 1990 [13]. As a result, certain
indicators used in earlier work had to be eliminated. At this
point, it is also appropriate to stress that we shall not be
presenting mere zero-order associations, but partial associations
between ideational dimensions and demographic choices after
controlling for age, education and socio-economic position. The
second issue concerning trends in ideational factors has only been
analyzed with EVS-data for Belgium only [14], and for the rest, the
EVS-data have been seriously underutilized. At present, we shall
extend the analysis to the four countries mentioned earlier,
thereby also taking advantage of a much larger pooled sample.
The trends by cohort in the 1981-1990 decade are of prime
interest for our basic research question: ®Are ideational
factors among the younger generations still markedly shifting in
the directions that were supportive of cohabitation or delayed
parenthood and low fertility? Or, alternatively, are the younger
generations in the four countries exhibiting trend reversals with
respect to basic value orientations that may forecast the end of
the "second demographic transition"? [15]¯
The questions addressed in the current paper fall squarely in
the research traditions dealing with selection effects stemming
from ideational orientations and with the recursive models, as
described by points b and c in the introduction. The omission of
factors related to human capital and economic opportunity
structures, as described in point a of the introduction, should not
be mistaken for a deliberate neglect on our part. Quite the
opposite holds: we are convinced that such economic factors will
continue to codetermine the demographic choices in the individuals'
life course in the future.
With this caveat in mind, we shall now turn to the
identification of major ideational correlates of demographic
choices.
3. The ideational dimensions: four components of
conventionalism
In the EVS-data for 1981 and 1990 fourty items were selected
that had remained identical with respect to wording and response
categories. These fourty items were initially brought together in
14 subscales [16].
Then, the 14 subscales were further reduced to four
uncorrelated basic dimensions by means of a principle component
analysis. The statistical results are shown in tab. 1 in which we
have reported the correlations (i.e. the factor loadings) between
the four factors on the one hand and the 14 subscales together with
the 40 items on the other [17].
The four basic dimensions of conventionalism used
throughout this paper can be described as follows:
See file 7thtab1.asc p. 1
See file 7thtab1.asc p. 2
See file 7thtab1.asc p. 3
Factor I: religiosity and morality
This factor is predominantly identified by the first five
subscales and twenty items. Positive values correspond with high
levels of religiosity (belief in God and in sin, strength and
comfort from faith) and regular service attendance. This is
correlated with strict ethics concerning the interference with
matters related to life and death (abortion, suicide, euthanasia),
and with the accentuation of strict civil and sexual morality [18].
Factor II: traditional family orientation
This factor measures conventionalism with respect to family
composition (aversion to single parenthood), authority in the
family, mutual duties between parents and children, and the need
for children for life fulfilment. The factor is based on three
subscales and seven items.
Factor III: the rightist political orientation
The factor identifies aspects associated with the ®new
right¯, such as intolerance toward ethnic minorities, and aspects
of conventionalism with respect to labour relations and economic
control, such as the refutation of managerial involvement of
employees. Evidently, the indicators of the ®new left¯, such as a
preference for a simpler and more natural life style and for
reduced emphasis on money and material possessions, are negative
correlates of factor III. The factor is based on four subscales
and nine items.
Factor IV: Inglehart's materialism
The factor is strongly correlated with the ®materialist¯-end
of Inglehart's (1970, 1990) post-materialism scale, which is itself
based on Maslow's needs hierarchy (1954). In this respect,
®materialism¯ is not to be understood as an orientation toward
luxury or conspicuous consumption, but as a preoccupation with
basic economic and physical security. Hence, a preference in
favour of inflation control and the maintenance of law and order.
Post- materialists on the other hand are oriented to individual
self-actualization, imagination and creativity, and grass-root
democracy. They are well represented among the voters for the
European ®Green¯-parties. Factor IV is therefore also negatively
correlated with the item of more emphasis on personal development.
Factor IV is based on two subscales and four items. [19]
At this point it should be stressed that a number of other
basic value dimensions are missing, largely as a result of
incomparability of the scales used in the EVS 1981 and 1990 rounds.
As a consequence, we are unable to introduce a dimension concerning
female orientations toward labour force participation or career
development, a dimension dealing with the symmetry of partner
relations, or a dimension focussing on social responsibilities and
community service. However, many of the items related to such
issues were used in our earlier work [20].
4. The components of conventionalism and demographic
characteristics of the life course
The dependent variables in the present analysis are
current status characteristics corresponding to various steps in
the life- course. No retrospective questions about earlier states
and transitions were asked in the EVS-surveys. For the respondents
aged 20-29, the following dichotomous dependent variables were
constructed:
a. Home-leaving (1) versus home-staying (0) This variable is
only constructed for respondents without a sexual partner. Those
who have left the parental home may be living on their own or they
may be sharing quarters with room-mates.
b. Cohabiting (1) versus married (0) This variable is
constructed for all those with a partner. The failure of
including a simple retrospective question (ever cohabited?) in the
EVS results in the blurring of the distinction between those who
never and those who ever cohabited among those currently married.
Such a finer distinction would undoubtedly have produced larger
contrasts in value orientations between those who married directly
and those passing through cohabitation first than we can document
at present.
c. Parent (1) versus being childless (0) This distinction is
made for all respondents 20-29 who are at present in a union, i.e.
married or cohabiting. This variable is obviously an indicator of
earlier versus delayed parenthood.
For the age group 30-49, we have equally retained three
dependent variables:
a. Cohabiting (1) versus married (0) The dichotomy is set up
for respondents in a union only. It captures those with a
prolonged cohabitation experience beyond age 30, and also a small
group of divorced persons who have started a new ®paperless
marriage¯, as opposed to those who moved into a new marriage.
b. Still being childless (1) versus being a parent (0)
This contrast is established for all respondents currently in
a union. The variable identifies those with strongly delayed
parenthood, i.e. until at least age 30.
c. Progression to parity 3 or beyond (1) versus parities 1 or
2 only (0)
This dochotomy is being constructed only for the respondents
who are currently living in a union and who are parents.
Since value orientations are not independent of age,
education and socio- economic position, and since the latter
variables also have a direct effect on the demographic events in
the life-course, the association between a) the four dimensions of
conventionalism and b) the demographic dependent variables have all
been measured after controlling for age and the two socio-economic
correlates. The latter two control variables were measured as
follows:
a. Education It is the age at which full-time education was
terminated. For students, the age at which they expect the end of
schooling is entered. In the present analysis, this variable was
introduced by means of 4 categories ranging from termination prior
to age 15, 15-17, 18-20, and 21+. No information concerning the
type of schooling is present in the EVS.
b. Socio-economic position It is also a categorical variable
introducing the distinction between a) upper and middle level white
collar, and employers, b) lower white collar, self-employed, and
blue collar workers, c) students, d) unemployed, and e) for women,
housewives. The current socio-economic position categories help to
remedy the shortcomings of the education variable, whilst being
proxies for social status and individual income. Regretably, the
EVS does not contain information on the partner's education or
occupation, so that further refinements were not possible.
The results of the analysis are presented in tab. 2 through
4. Since the dependent variables are all dichotomies, logit
regression was utilized. The results come in the form of relative
odds ratios, indicating the likelihood of a specified demographic
event for a category of the ideational factors relative to a
reference category (odds ratio = 1.00). The four dimensions of
conventionalism were also categorized on the basis of quartiles.
Note, however, that the distribution on which the quartiles are
based is the distribution for the entire population in 1990,
irrespective of age (in earlier analyses, quartiles were age group
specific). One will also notice that the reference groups in tab.
1 through 4 are always those with the lowest scores on the four
dimensions of conventionalism. Since we are dealing with quartiles
for the entire population, the lowest score quartile (Q1) generally
has the highest number of cases if analyses are restricted to
younger cohorts. Finally, aside from the relative likelihood, also
the R-statistic of the partial contribution of the explanatory
variables is given, together with significance levels. These
partial contributions are those over and above the contributions
provided by the three control variables (age, education, socio-
economic position). We shall now turn to the results.
Among the 20-29 year old women (see tab. 2), home leaving is
strongly related to factors II and III. The probabilities of home
leaving diminish markedly as the respondents hold traditional
family values and conventional views with respect to material life
style, authority in work, economic structure and ethnic relations.
The partial contribution of the other two factors in the 7-variable
model is negligible, but the relative odds ratio are again in line
with earlier findings: home leaving diminishes with increasing
religiosity and stricter ethics and with the progressive moving
away from Inglehart's ®post-materialism¯.
For the same age groups of respondents, the likelihood of
cohabitation rather than marriage is strongly related to factors I
and II. The negative relationship between cohabitation and
religiosity (factor I) is a classic, and still very prominent in
1990 in the four countries concerned. Also the relationship with
factor II is as expected: the likelihood of cohabitation declines
markedly as one moves away from the least conventional family
orientations. Factor III contributes little to the model, but the
relative odds ratios are still in line with our expectations: there
is evidence that cohabitation is most prominent among young women
in a union who hold political views associated with the ®new left¯,
such as greater tolerance for ethnic minorities, rejection of full
owner and managerial authority, and accentuation of a ®simpler and
more natural¯ life style. With respect to cohabitation, only
factor IV produces no contribution and no patterning either.
Earlier progression to parenthood for women 20-29 in a union
is strongly related to factors II and IV, but the patterning for
factor I is equally in line with our expectations. First,
orientations toward traditional family values and intergenerational
duties are strong net predictors of earlier parenthood (factor
III), whereas those who are in the post- materialist quartile
accentuating individual self-actualization have by far the lowest
probability of already being a parent (factor IV). Next in line is
the classic finding that earlier parenthood is positively
associated with religiosity (factor I). The contribution of factor
III and its political dimension is zero, and the patterning is also
not as expected.
Hence, for younger women, many differentiations for the value
dimensions do contribute significantly to the explanatory power of
the model, and all patternings except one are perfectly in line
with what is expected. See file 7thtab2.doc p. 1 See file
7thtab2.doc p. 2 See file 7thtab3.doc p. 1 See file 7thtab3.doc
p. 2 The results for the men 20-29 are presented in tab. 3. The
general story is not very different from the one just told. Home
leaving is again very clearly negatively associated with factor II,
being conventionalism in family orientations and intergenerational
duties. Also factor I yields no surprise: it contributes
significantly to the model as well, and the patterning of relative
risks shows the negative net association between home leaving and
religiosity or strict ethics. Factor III contributes nothing for
the younger male respondents, in contrast to what was found for
younger women, but the patterning of the relative odds ratios is
similar for the two sexes. Only factor IV, Ingleharts
®materialism¯, could be dropped.
The position of cohabitants among younger men with a partner,
as opposed to being married, is strongly and negatively associated
with factors I and II, being religiosity and conventionalism in
family relations. These were also the two dominant value
dimensions in the analysis for women. The other two factors do not
contribute to the model, but the patterning of the relative odds
ratios is convincing: cohabitation is also negatively related to
the conventional political orientations and positively to
Inglehart's ®post- materialism¯. The same general line emerges
with respect to parenthood. A very dominant positive relationship
emerges with conventional family orientation (factor II). Next is
a significant contribution by factor IV, showing a strong negative
association between parenthood and ®post-materialism¯. The classic
positive association between earlier parenthood and religiosity is,
however, less pronounced among men, which is mainly due to the odd
outcome for the second quartile. Factor III, finally contributes
nothing to the overall explanatory power, and the patterning of the
relative odds ratios is not concordant with expectations.
Paternity, according to these results, tends to be negatively
associated with the adherence to the ®new right¯. The finding for
women went somewhat in the same direction, and also constituted the
only exception.
The analysis for women aged 30-49 is presented in tab. 4.
Most such women have obviously moved on to marriage, but the
likelihood of still being in a cohabiting union or of having
returned to such a union is significantly and negatively related to
conventional family values (factor II). It is also clearly
negatively related to religiosity (factor I). The other two
factors produce no significant results, nor a convincing patterning
of relative odds ratios. Similarly, most women aged 30-49 and
living in a union have had their first child. Postponed maternity
is again negatively associated with the first two factors, being
conventional family values and religiosity or strict ethics.
Factors III and IV produce no contribution to the model, but the
unexpected result with respect to greater childlessness for the
highest quartile (Q4) of the ®new right¯ orientation shows up
again.
Progression to parities 3 or higher are a more rare event.
The likelihood of this occurring among women 30-49 who are already
parents, is strongly related to factors II (conventional family
orientation), I (religiosity and strict ethics) and IV (Inglehart's
®materialism¯). In all three cases the patterning of relative odds
ratios is the classic one. Only factor III (®new right¯) is
unnecessary in the prediction, and fails to produce a clear
patterning as well.
On the whole, the contributions of the four ideational
dimensions for women aged 30-49 are smaller than for the younger
women, but the directions of the associations are again the
expected ones. There is again only one exception pertaining to
factor III, with respect to those in the highest quartile of the
®new right¯ being significantly more likely still to be childless
than those in the lowest quartile.
The last set of statistical results, i.e. for men aged 30-49,
are presented in tab. 5. The outcomes are again similar to those
for women. In fact, the likelihood to be in a cohabiting rather
than a married partnership is more strongly related to conventional
family orientations (factor II) and religiosity or strict ethics
(factor I) for men than for women. Also factor III exhibits a
regular pattern, with those in the highest quartile with respect to
the ®new right¯ showing the least inclination to cohabitation.
The feature of not yet being a father is the more difficult
one to predict on the basis of the four value dimensions used here.
Just as for the women in the same age group, only the contribution
of factor II is significant, but not surprising. Further more,
among men, the probability of still being childless only seems to
be smaller among the highest quartile with respect to religiosity,
but on the whole, factor I contributes nothing to the prediction.
Perhaps the only surprising feature, as already repeatedly
mentioned, is again the tendency that those with more ®new right¯
orientations are more likely still to be childless. This feature
of lower fertility among men with a greater preoccupation with
material comfort and with strict lines of authority in business, an
aversion to leftist experiments, and a distrust for ethnic
minorities, is also emerging in the significantly lower proportions
of them moving to parity three or higher. It looks as if old
securities in
See file 7thtab4.asc p. 1
See file 7thtab4.asc p. 2
See file 7thtab5.asc p. 1
See file 7thtab5.asc p. 2 terms of standards of living and
lines of authority need to be defended, resulting in an aversion to
new-comers (i.e. immigrants) and also in restricted fertility.
We can conclude this section with the observation that,
independently from age, education and socio-economic position, the
value dimensions are again often showing robust associations with
the demographic aspects of the life course. Particularly factors
I and II, being the classics of religiosity and conventional
familism, are producing significant and consistent outcomes. The
main surprise - already signaled in an earlier paper (Lesthaeghe
and Moors, 1994) - with the EVS data for the four countries is that
high scores on the ®new right¯ dimension tend to be associated with
lower fertility, particularly among men. The interpretation given
above to this finding is obviously only tentative and the feature
itself requires confirmation by other data sets.
Having established that value dimensions are by no means
redundant variables in the study of the demographics of the life-
course, we shall now move on to the main section of the paper, and
focus on trends in these dimensions of conventionalism.
5. The trends in the ideational components by cohort,
1980-1990
Repeated measurement over several decades would allow us to
inspect whether or not birth cohorts would maintain their
respective opinions over time. This would be an indication that
the cohort-model, proposed by socialization theory, is indeed
preponderant. The alternatives are that cohorts would simply
follow in the footsteps of their predecessors once they reach the
same age (life-cycle effect), or that all cohorts simultaneously
shift positions at the same time, irrespective of their age (period
effect).
In this paper we only have two measurements that are nine
years apart, which is inadequate to make firm inferences about the
respective dominance of cohort, life-cycle or period effects
respectively. The most we can do is to present a picture of the
trend for the 1980s, and indicate whether or not cohorts have
maintained their positions. In addition, it is also interesting to
see where the youngest cohort is located in the 1990 surveys
relative to their predecessors. This may give us an idea of future
trends to expect in the ideational components that codetermine
demographic outcomes.
The obvious way to do this is to plot the outcomes for each
of the factors or their subscales (vertical axis) by cohort
(horizontal axis) for the two dates of measurement [21]. The
orthogonal factors have a mean of zero and a standard deviation of
unity. The subscales are based on summated dichotomous ratings,
but comparability between them in terms of scale is restored by
dividing these summated ratings by the number of items. The
outcomes then vary between 0 and 1. We shall now inspect the
results.
a. Trends in religiosity and morality (factor I)
The scores on factor I are plotted on fig. 1. The oldest
cohorts born prior to 1920 still had average scores of more than
half a standard deviation above the mean, whereas the younger ones,
born after 1950 are nearly half a standard deviation below the
mean. On the whole, however, the measurements for 1990 by cohort
are almost identical to those for 1981, showing that cohort
stability was present at least during the last decade. The
position of the youngest generation, born after 1960, also
indicated that there may be a slight attenuation in the overall
downward profile. This attenuation is due to two subscales: the
youngest cohort is slightly less likely to accept abortion than its
two predecessors (subscale GABORT), and shows no further decline in
religiosity (subscale RELIG) compared to the cohort born in the
1950s (see fig. 2). The two subscales for respectively civil and
sexual morality, on the other hand, show no such attenuation: the
most recent cohort born in the 1960s adds scores that are, on
average, slightly lower than that of their predecessors. The
overall picture with respect to factor I is hence characterized for
the 1980s by a lack of shifts of cohort averages [22] and by a
smaller change contributed by the youngest cohort, compared to the
large changes that earlier cohorts made relative to the position of
their immediate predecessors. However, it would be an exaggeration
to speak of a systematic trend reversal for the youngest cohort,
since not all subscales exhibit the levelling-off feature.
See file 7thfigur.xls fig. 1
See file 7thfigur.xls fig. 2
b. Trends in traditional family orientation (factor II)
As shown in fig. 3, the contrasts between the youngest and
oldest cohort are equally of the order of one standard deviation,
with the youngest cohort exhibiting the least adherence to
traditional family norms. However, in the decade 1980-1990, there
may have been a return to firmer family values for almost all
cohorts concerned: the 1990 measurements tend to be located above
the 1981 ones. Such a feature is indicative of a systematic period
effect in which all cohorts shift positions in the same direction.
The outcome for the youngest cohort is that it is located in the
same place as its immediate predecessor about ten years earlier.
The upward shift of all cohort-points is contributed by two
subscales, as shown in fig. 4. The opinions in favour of a
traditional family composition (both parents needed, children as
life-fulfilment in subscale TRAGEZ), and the opinions concerning
the mutual duties of parents and children (subscale PCREL) have
been asserted to a greater extent in 1990 than in 1981 by all
cohorts, resulting in the fact that the point for the latest cohort
in 1990 is now located above that of its predecessor in 1981. For
the third subscale, reflecting an opinion that puts more emphasis
on authority and the family, the shift is much less pronounced but
still present for the cohorts born after 1940, and equally results
in the feature that the youngest cohort born after 1960 has made no
changes compared to the cohort born in the 1950s as measured in
1981.
The overall outcome for factor II is that all cohorts tended
to move to more traditional family values during the 1980s, and
that the youngest cohort is located where its two predecessors were
in 1981. c. Trends in rightist political orientation (factor III)
Fig. 5 shows the overall shift upward with respect to the
dimension measured by factor III: all cohorts have again moved to
more political conservatism during the 1980s, and the shift is
quite remarkable. The only feature that is intact is the general
downward slope with age, except for the position of the latest
cohort born in the 1960s.
The overall upward shift is also found in almost all
subscales that constitute factor III (see fig. 6). First, all
cohorts return to the more authoritarian forms of management with
less interference from employees and firmer control of owners
(subscale MANAG). Second, all cohorts have become more aversive to
extremists (subscale EXTREM). Third, all cohorts are slightly more
likely not to want
See file 7thfigur.xls fig. 3
See file 7thfigur.xls fig. 4
immigrants as neighbours (subscale INTOL). Fourth, there is
no enhanced assertion in the direction of a ®more single and more
natural life¯ and no reduced emphasis on money and possessions
(subscale NLEFT). With respect to the latter two subscales, there
is hardly any cohort differentiation at both times of measurement.
Hence, the changes in factor III show that there is a uniform
and general return to top- down management principles and a scare
for radicals and immigrants alike. The dream in favour of a ®more
natural life style¯ is over, and has probably faded away in the
memory of the youngest generation [23].
d. Trends in Inglehart's materialism (factor IV)
The general feature with respect to this factor is again that
of a change which has affected all cohorts during the 1980s: the
proportions preoccupied simultaneously with inflation and
maintaining order have diminished throughout (see fig. 7). In
fact, the proportions at the ®materialist¯-end of the Inglehart-
scale are nearly undifferentiated at both times of measurement in
all cohorts born since the 1930s. Only the older generations still
have a more marked ®pure materialist¯ profile.
The decline in ®pure materialism¯ during the 1980s is likely
to be associated with the fact that inflation rates diminished in
the four countries concerned, and are at present hardly a
preoccupation of the general public. Employment, by contrast, has
been of much greater concern, but this issue is not reflected in
the Inglehart scale. Finally, the other subscale correlated with
factor IV (but negatively), and measuring more emphasis on personal
development and on technology, shows essentially a flat profile
across cohorts and hardly any shift in the 1981-1990 period. Only
the two oldest cohorts, born prior to 1920, put less emphasis on
these two issues.
See file 7thfigur.xls fig. 5
See file 7thfigur.xls fig. 6
See file 7thfigur.xls fig. 7
6. Conclusions
In the light of other research, the findings with respect to
the first two ideational dimensions contain no surprises:
a. Religiosity, civil and sexual morality, and traditional
family orientations are negatively associated with independent
living and with cohabitation, and they are positively related to
earlier parenthood and progression beyond two children.
b. These ideational dimensions exhibit strong differences
between cohorts: the younger the cohort, the weaker the adherence
to these value orientations becomes.
c. At least in the period 1981-1990, the various cohorts
maintained their positions with respect to religiosity and morality
issues, or they all moved slightly in the direction of stronger
familistic orientations.
d. The most recent cohort, born in the period 1961-1972, has
so far not continued the trend toward greater secularization or
less familism. But, there is no indication of a definite trend
reversal either. Obviously, data for the cohort born in the 1970s
are needed to determine whether this waring off in the cohort
profile corresponds with a temporary plateau only, or with a
genuine trough in a U-shaped long term development.
The findings with respect to the other ideational factors are
more complex:
a. The factor ®political right¯, with its references to
conventionalism in management, authority, ethnic relations and
consumerism, was negatively associated with independent living and
cohabitation. This was expected, given that new patterns in
partner relationships had been pioneered by the ®new left¯ during
the 1960s and 1970s in the four countries concerned.
b. But, quite unexpectedly, this factor was positively
related to postponed parenthood, and, among men, also to less
progression to larger family sizes.
c. Inglehart's ®materialism¯ did not behave as predicted with
respect to home leaving and cohabitation, but proved to be a
significant predictor of earlier parenthood and, among women, also
of progression beyond two children.
d. Factors III and IV exhibit marked shifts for all cohorts,
in the direction of more politically rightist orientations, more
consumerism, and less ®pure materialism¯ in the Maslowian sense.
Both in 1981 and 1990, the various cohorts born after 1930 showed
little differentiation between themselves. Hence, these two
ideational factors proved to be particularly sensitive to period
fluctuations and insensitive to cohort patterning.
Could we, on the basis of these findings, now conclude that
the youngest cohorts in the four countries are exhibiting decisive
ideational trend reversals? Our conclusion is that it is
definitely too early to come up with such a prediction. At
present, all that can be said is that the ideational props that
have sustained the second demographic transition seem to have lost
their momentum during the 1980s, and that the latest cohort born in
the 1960s is almost completely undifferentiated - at least on the
items used here - from their immediate predecessors as measured in
1990.
Can this finding be taken as a prediction for declining
proportions cohabiting, earlier marriage and parenthood, less
divorce and more progression to higher parities? We think not.
First, the ideational components related to religiosity, civil and
sexual morality and traditional family relations have not yet
exhibited a genuine trend reversal, only a loss of momentum.
Second, the patterns of family formation are also codetermined by
economic factors, as outlined in the introduction. These economic
determinants have not been considered here, and, hence, a paper
that is complementary to this one is needed to trace the changes in
female economic autonomy, opportunity costs, and prospects for
employment and income for the new cohorts. We are, furthermore,
not sure that prospective developments with respect to these issues
would be supportive of the ®coming back of the old family¯. More
likely is that various forms of family formation will continue to
coexist, and that the rapid growth period of less conventional
family patterns may have come to an end. In short, diversity is
likely to prevail in the next decade, but the relative shares of
each type may not be changing all that much any more. ®Stability
in diversity¯ seems to be the more appropriate description for the
near future.
7. References
Avery R., F. Goldscheider, A. Speare (1992), Feathered
Nest/Guilded Cage: Parental Income and Leaving Home in the
Transition to Adulthood, Demography, 29, 375-388.
Axinn W., A. Thornton (1992), The Influence of Parental
Resources on the Timing of the Transition to Marriage, Social
Science Research, 21, 261-285.
Axinn W., A. Thornton (1993), Mothers, Children and
Cohabitation - The Intergenerational Effect of Attitudes and
Behavior, American Sociological Review, 58, 233-246.
Beck P., M. Jennings (1975), Parents as Middlepersons in
Political Socialization, The Journal of Politics, 37, 83-107.
Becker G. (1981), A Treatise on the Family, Harvard UP,
Cambridge Mass.
Bumpass L. (1990), What's Happening to the Family?
Interactions between Demographic and Institutional Changes,
Demography, 27 (4), 483-498.
Clarkberg M., R. M. Stolzenberg, L. J. Waite (1993),
Attitudes, Values and the Entrance into Cohabitational Unions,
Research report, Population Research Center, Norc and the
University of Chicago, Chicago IL.
Grimmins E., R Easterlin., Y. Saito (1991), Preference Change
among American Youth: Family, Work and Goods Aspirations,
1978-1986, Population and Development Review, 17, 115-133.
de Jong-Gierveld J., A. Liefbroer, E. Beekink (1991), The
Effect of Parental Resources on Patterns of Leaving Home among
Young Adults in the Netherlands, European Sociological Review, 7,
55-71
de Feyter H. (1991), Voorlopers bij demografische
veranderingen, Nidi-report 22, Netherlands Interdisciplinary
Demographic Institute, The Hague.
Easterlin R. (1976), The Conflict between Aspirations and
Resources, Population and Development Review, 2 (3), 417-426.
Easterlin R., C. Macdonald, D. J. Macunovich (1990), How Have
American Baby Boomers Fared? Earnings and Economic
Well-being of Young Adults 1964-1987, Journal of Population
Economics, 3 (4), 277-290.
Easterlin R., E. Crimmins (1991), Private Materialism,
Personal Self-fulfilment, Family Life and Public Interest: the
Nature, Effects and Causes of Recent Changes in the Values of
American Youth, Public Opinion Quarterly, 55, 499-533.
Ekert-Jaff‚ O., C. Sofer (1991), Un point de vue
d'‚conomiste sur la nuptialit‚, in T. Hibert, L. Roussel (eds.), La
nuptialit‚ - Evolution r‚cente en France et dans les pays
d‚velopp‚s, Ined CongrŠs et Colloques 7, Presses Universitaires de
France, Paris, 19-35.
England P., G. Farkas (1986), Households, Employment and
Gender - A Social, Economic and Demographic View, Aldine De
Gruyter, New York.
Gecas V., M. Seff (1990), Families and Adolescents - A Review
of the 1980s, Journal of Marriage and the Family, 52, 941-958.
Goldscheider F., L. Waite (1991), New Families, no
Families? The Transformation of the American Home, University of
California Press, Berkeley.
Goldscheider F., C. Goldscheider (1993), Leaving Home before
Marriage - Ethnicity, Familism and Generational
Relationships, University of Wisconsin Press, Madison.
Goldscheider F., C. Goldscheider (1994), Leaving and
Returning Home in 20th Century America, Population Bulletin,
Population Reference Bureau, 48 (4), 1-34.
Grossbard-Schechtman S. (1993), On the Economics of
Marriage - A Theory of Marriage, Labor and Divorce, Westview Press,
Boulder.
Hoem J. (1986), The Impact of Education on Modern
Family-union Initiation, European Journal of Population, 2,
113-133.
Hoem B., J. Hoem (1989), The Impact of Women's Employment on
Second and Third Births in Modern Sweden, Population Studies, 43,
47-67.
Inglehart R. (1970), The Silent Revolution, Princeton UP,
Princeton N.J.
Inglehart R. (1985), Aggregate Stability and Individual Flux
- The Level of Analysis Paradox, American Political Science Review,
79, 97-116.
Inglehart R. (1990), Culture Shift in Advanced Industrial
Society, Princeton UP, Princeton N.J.
Kiernan K. (1992), The Impact of Family Disruption in
Childhood on Transitions Made in Young Adult Life, Population
Studies, 46, 213-234.
Kohn M. L. (1977), Class and Conformity - A Study in
Values, University of Chicago Press, Chicago, Ill.
Liefbroer A. (1991), Choosing between a Married and an
Unmarried First Union among Young Adults: a Competing Risk
Analysis, European Journal of Population, 7, 273-298.
Lesthaeghe R., D. Meekers (1986), Value Changes and the
Dimensions of Familism in the European Community, European Journal
of Population, 2, 225-268.
Lesthaeghe R., D. van de Kaa (1986), Twee demografische
transities? in R. Lesthaeghe, D. van de Kaa (eds.), Groei of Krimp?
van Loghum Slaterus, Deventer, 9-24.
Lesthaeghe R., G. Verleye (1992), De tweede demografische
transitie - Conceptuele basis en recente evolutie, in J. de Jong-
Gierveld, N. van Nimwegen (eds.), De demografische uitdaging -
Nederland in Europa op weg naar de 21e eeuw, Bohn, Stafleu, Van
Loghum Uitgevers, Houten and Zaventem, 15-49.
Lesthaeghe R., J. Surkyn (1988), Cultural Dynamics and
Economics of Fertility Change, Population and Development Review,
14, 1-45.
Lesthaeghe R., G. Moors (1992), De gezinsrelaties: de
ontwikkeling en stabilisatie van patronen, in J. Kerkhofs, K.
Dobbelaere, L. Voy‚, B. Bawin (eds.), De versnelde ommekeer - De
waarden van Vlamingen, Walen en Brusselaars in de jaren negentig,
Uitgeverij Lannoo, Tielt, 19-68.
Lesthaeghe R., G. Moors (1994), Expliquer la diversit‚ des
formes familiales et domestiques: th‚ories ‚conomiques et
dimensions culturelles, Population, 6, 1503-1526.
Lesthaeghe R., G. Moors (1995), Living Arrangements,
Socio-economic Position and Values among Young Adults: a Pattern
Description for France, Germany, Belgium and the Netherlands 1990,
forthcoming in D. Coleman (ed.), Europe's population in the 1990s,
Oxford UP, Oxford (chapter 6).
Leridon H., C. Villeneuve-Gokalp (1988), Les nouveaux couples
- Nombre, caract‚ristiques, attitudes, Population, 43 (2), 331-374.
Lye D. N., I. Waldron (1993), Correlates of Attitudes toward
Cohabitation, Family and Gender Roles, Seattle Population Research
Center Working paper, 93-10, University of Washington and Battelle,
Seattle WA.
Manting D. (1994), Dynamics in Marriage and Cohabitation -An
Inter-temporal Life Course Analysis of First Union Formation and
Dissolution, Ph.D.-dissertation Universiteit Amsterdam,
Planologisch and Demografisch Instituut, Amsterdam.
Maslow A. (1954), Motivation and Personality, Harper and Row,
New York.
Mc Lanahan S., G. Sandefur (1994), Growing up with a
Single Parent, Harvard UP, Cambridge Mass.
Michael R. T., N. Tuma (1995), Entry into Marriage and
Parenthood by Young Men and Women: the Influence of Family
Background, Demography, 22, 515-544.
Miller W. B. (1992), Personality Traits and Developmental
Experiences as Antecedents of Childbearing Motivation, Demography,
29 (2), 265-285.
Moors G. (1995), Gezinsvorming en processen van
waardenselectie en -aanpassing, Steunpunt voor Demografie Working
Paper, 95-2, Vrije Universiteit Brussel, Brussels.
Oppenheimer V. (1988), A Theory of Marriage Timing,
American Journal of Sociology, 94, 563-591.
Peterson G. W., B. C. Rollins (1987), Parent- child
Socialization, in B. Sussman, S. Steinmetz (eds.), Handbook of
Marriage and the Family, Plenum Press, New York, 471-507.
Pollack R. (1976), Interdependent Preferences, American
Economic Review, 66, 309-320.
Pollack R. (1985), A Transaction Cost Approach to Families
and Households, Journal of Economic Literature, 23, 581-608.
Pollack R., S. Watkins (1993), Cultural and Economic
Approaches to Fertility: a Proper Marriage or a M‚salliance?,
Population and Development Review, 19, 467-496.
Rezsohasy R. (1991), Les nouveaux enfants d'Adam et Eve -Les
formes actuelles des couples et des familles, Academia, Louvain-
la-Neuve.
Ryder N. (1965), The Cohort as a Concept in the Study of
Social Change, American Sociological Review, 30, 843-861.
Santow G., M. Bracher (1994), Change and Continuity in the
Formation of First Marital Unions in Australia, Population
Studies, 48, 475-496.
Thave S. (1991), C‚libataires mais pas seuls - Evolution
r‚cente des cohabitations de c‚libataires, in T. Hibert, L. Roussel
(sous la dir.), La nuptialit‚ - Evolution r‚cente en France et dans
les pays d‚velopp‚s, Ined CongrŠs et Colloques, 7, Presses
Universitaires de France, Paris, 59-74.
Thornton A., D. F. Alwin, D.Camburn (1983), Causes and
Consequences of Sex-role Attitudes and Attitude Change, American
Sociological Review, 48, 211-227.
Thornton A., D. Camburn (1987), The Influence of the
Family on Premarital Sexual Attitudes and Behavior, Demography, 25
(3), 323-340.
Thornton A. (1989), Influence of the Marital History of
Parents on the Union Formation Experience of Children, unpublished
paper, Institute for Social Research and Population Studies Center,
University of Michigan, Ann Arbor, MI.
Thornton A. (1991), Influence of the Marital History of
Parents on the Marital and Cohabitational Experiences of Children.
American Journal of Sociology, 96, 868-894.
Thornton A., W. G. Axinn, D. H. Hill (1992), Reciprocal
Effects of Religiosity, Cohabitation and Marriage, American Journal
of Sociology, 98 (3), 628-651.
Thornton A. (1992), The Influence of the Parental Family on
the Attitudes and Behavior of Children, in S. J. South, S. E.
Tolnay (eds.), The Changing American Family: Sociological and
Demographic Perspectives, Westview Press, Boulder CO, 247-266.
Trost J. (1978), Attitudes to and Occurrences of
Cohabitation and Marriage in Sweden, Journal of Divorce, 2,
415-421.
van de Kaa D. (1987), Europe's Second Demographic
Transition, Population Bulletin, Population Reference Bureau, 42
(1), 1-58.
Villeneuve-Gokalp C. (1990), Du mariage aux unions sans
papiers: histoire r‚cente des transformations conjugales,
Population, 45 (2), 265-298.
Weinstein M., A. Thornton (1989), Mother- child Relations and
Adolescent Sexual Attitudes and Behavior, Demography, 26 (4),
563-577.
Wu L. L., C. Martinson (1991), Family Structure and the Risk
of a Premarital Birth, Nsfh-Working paper, 45, Center for
Demography and Ecology, University of Wisconsin, Madison WI.
_______________________________
1. For human capital effects and especially the role of
female economic autonomy, see G. Becker, 1981; J. Hoem, 1986; B.
Hoem and J. Hoem, 1989; O. Ekert- Jaff‚ and C. Sofer, 1991; S.
Grossbard- Schechtman, 1993. The economic relative deprivation
theory is obviously well represented in the writings of R.
Easterlin and colleagues, e.g.: R. Easterlin, 1976; Easterlin et
al., 1990; E. Crimmins et al., 1991.
2. V. Oppenheimer (1988) connects the rise of premarital
cohabitation and postponement of marriage to increased quality
aspirations, especially among better educated women, and to the
longer search for a ®minimally acceptable match¯.
3. For instance: R. Avery et al., 1992; de Jong-Gierveld et
al., 1991; W. Axinn and A. Thornton, 1992.
4. See for instance R.T. Micheal and N. Tuma, 1985; A.
Thornton, 1991; A. Thornton and D. Camburn, 1987; K. Kiernan, 1992;
S. McLanahan and S. Sandefur, 1994.
5. E.g. J. Trost, 1978; H. Leridon and C. Villeneuve-Gokalp,
1988; C. Villeneuve- Gokalp, 1988; H. de Feyter, 1991; S. Thave,
1991.
6. For instance P. Beck and M. Jennings, 1975, and M.L. Kohn,
1977 for selection effects with respect to political orientations;
R. Lesthaeghe and J. Surkyn, 1988, or W. Axinn and A. Thornton,
1993, for selection effects on family building patterns.
7. See, for instance, R. Lesthaeghe and D. Meekers, 1986;
Thornton et al., 1992; C.Villeneuve-Gokalp, 1990; L. Bumpass, 1990;
A. Liefbroer, 1991; F. Goldscheider and L. Waite, 1991; M.
Clarkberg et al., 1993; D. Lye and I. Waldron, 1993; G. Santow and
M. Bracher, 1994; D. Manting, 1994; R. Lesthaeghe and G. Moors,
1994; F. Goldscheider and C. Goldscheider, 1993 and 1994.
8. See N. Ryder, 1965. Social metabolism refers to the fact
that changes in values in the society as a whole stem from the fact
that older values die off with the older generations, whereas new
ones are generated by new cohorts. A well known example of such a
model is found in Inglehart's cohort patterning of the rise of
post-materialism. See R. Inglehart, 1970, and especially 1990.
9. See for instance, A. Thornton et al., 1992, for results of
the Detroit panel, or G. Moors, 1995, for findings from the
Bielefeld panel.
10. See G. Becker, 1981; P. England and G. Farkas, 1986; R.
Rezsohasy, 1991; R. Pollack, 1976 and 1985; R. Pollack and S.
Watkins, 1993.
11. See A. Thornton et al., 1992, for the example of older
cohorts changing their views on sexuality, gender relations and
cohabitation in function of the experience of their children.
12. See R. Lesthaeghe and G. Moors, 1994 and 1995.
13. The organizers changed many attitude batteries by adding
or dropping items, probably with the intention of improving the
instruments. In the process, however, they lost on comparability,
especially when picking or ranking procedures were involved. For
instance, the simple omission of ®religious faith¯ as a quality to
be stressed in educating children in the 1990-round resulted in the
incomparability of the entire battery, which had proved to be an
excellent determinant of the demographic variables. In the present
paper we can obviously only make use of questions and batteries
that are rigorously identical in the 1981 and 1990-rounds.
14. See R. Lesthaeghe and G. Moors, 1992.
15. On the notion of the ®second demographic transition¯ see
D. van de Kaa, 1987; R. Lesthaeghe and D. van de Kaa, 1986; R.
Lesthaeghe and G. Verleye, 1992.
16. The 14 subscales are:
A - The interference with life and death (INGLEV), i.e. the
respondents' attitudes regarding abortion, euthanasia and suicide;
B - Religiosity (RELIG), measured via the regularity of
service attendance, the belief in God and in sin, and the
procuring of comfort and strength from one's religious faith;
C - Acceptance of abortion (GABORT) which measures, via a
Guttman-scale, up till what level an induced abortion is
justified, i.e. never, only if the life of the mother is in danger,
if the child risks to be handicapped, if the mother is not married,
or also if no more children are desired;
D - Civil morality (CIVIL), based on the complete
refutation of five types of deviance, being accepting bribes,
claiming unentitled state benefits, free riding on public
transportation, tax evasion and keeping found money;
E - Sexual morality (SEXMOR), based on four indicators, being
the complete refutation of the notion of sexual freedom,
extramarital relations, prostitution and homosexuality;
F - Authority and family (NRIGHT) when respondents
prefered changes in the direction of more emphasis on authority and
on family life (as compared to less emphasis on money, material
possessions, or work; or greater emphasis on personal development,
on a more natural and simpler life style, or on the development of
technology);
G - Mutual duties of parents and children (PCREL),
measured through the choice for two sets of two alternatives.
These are: a) ®regardless of what the qualities and faults of one's
parents are, one must always love and respect them¯ (versus ®one
does not have the duty to respect and love parents who have not
earned it by their behaviour and attitudes¯); b) ®parents' duty is
to do their best for their children even at the expense of their
own well-being¯ (versus ®parents have a life of their own and
should not sacrifice their own well-being for the sake of their
children¯).
H - Traditional opinions concerning family composition
(TRAGEZ), based on the statements that a child needs both a mother
and a father, that single motherhood is not acceptable, and that a
woman needs children for her life fulfilment;
I - Intolerance toward ethnic minorities (INTOL), via the
items: immigrants or persons of another ethnic group not wanted as
neighbours;
J - Simpler life style (NLEFT), agreeing with: in the future
less stress on money and possessions and more emphasis on a simple
and more natural life style (see also subscale F);
K - Aversion political extremists (EXTREM), via not wanted as
neighbours: right wing extremists, left wing extremists;
L - Authority in labour relations (MANAG), with three items,
the first being ®the owners should run their business or appoint
the managers¯ (versus ®employees should participate in selection of
managers¯, ®the state should be the owner¯, ®the employers should
own the business and select managers¯); the second being that ®one
should always follow one's superiors' instructions¯ (versus ®must
be convinced first¯, ®depends¯); and the third being the choice in
a vignette- situation in favour of more pay according to
performance, even if job description and qualifications are
identical;
M - Inglehart materialism (MAT), with prefering the two
materialist items ®fighting rising prices¯ and ®maintaining order
in the nation¯ over the two post- materialist items (®giving people
more say in important decisions¯ and ®protection of freedom of
speech¯) or over a ®mixed¯ choice;
N - Stress on development technology and individual (ONTW),
agreeing with in the future more emphasis on the development of
technology and on the development of the individual (see also
subscales F and J).
17. The present principle component analysis differs from the
one presented earlier by the authors (see Lesthaeghe and Moors,
1994) in two respects. First, another set of indicators had to be
used here to maintain the comparability between the 1981 and 1990
measurements, and second, the present analysis is based on all
respondents irrespective of age (and not just on selected age
groups) in order to compare the various cohorts. The construction
of basic ideational dimensions for specific age groups is a finer
procedure, which had to be abandoned in the present paper for the
reasons mentioned above. A word should also be said about the
procedure adopted with respect to missing values on any of the
fourty items. The general procedure was to substitute the country,
age and gender specific mean. However, a factor score remained
missing if less than 10 of the 14 subscales were available.
18. It should be noted that the items concerning civic
morality were less strongly associated with religiosity among the
20-29 year old respondents in 1990 than in the sample as a whole.
19. See R. Lesthaeghe and D. Meekers (1986), R. Lesthaeghe
and G. Moors (1994, 1995). It should also be noted that a number
of value orientations such as those concerning the symmetry of
partner relations or the socialization values are strongly
affected by the influence of the partner on the presence of a
child, so that strong ex post rationalizations occur (see G. Moors,
1995). By contrast, religious outlook, political and ethical
convictions, xenophobia, notions of autonomy and authority in
general, and concerns about basic economic and physical
security are more remote value orientations with a firmer
anchoring in the socialization phase and in the religious and/or
social class values emphasized in the family of origin.
20. The meaning of factors II through IV was also
validated by checking the correlations with other items in the
1990- round. Good predictors of factor II proved to be a) having
children is a precondition for a successful marriage,b) the child
suffers if the mother works, c) hard work, thrift, and obedience
are qualities to be stressed in education. ®Imagination¯ is a
strong negative correlate of factor II. This pattern shows that
our interpretation of factor II is sound. Other correlates of
factor III were disapproval of the human rights and the
anti-apartheid movements, which are again indicators of the
political right. Finally, the additional correlates of factor IV
are the disapproval of the ecology movement and the other
indicators of ®materialism¯ in Inglehart's scale, i.e. ®a stable
economy¯ and ®the fight against crime¯. All the items in the
battery of self- esteem (feeling successful, able to convince
others, good at getting what I want, envied by others, rarely
unsure, giver of advice) are negative correlates of factor IV.
21. The sample sizes are as follows: 5466 for 1981 and 6234
for 1990. The smallest cohort contained 555 respondents (cohort
born prior to 1910, 1981 surveys) and the largest 1590 respondents
(cohort born in 1961-72, in the 1990 surveys).
22. This does of course not preclude individual flux
within a given cohort. It only means that such shifts in
individual positions of members of the same cohort are mutually
offsetting (cf. R. Inglehart, 1985).
23. This finding is in line with that of E. Crimmins et al.
(1991) for the American youths in the period 1978-1986. In these
data, career and especially money and material possessions have
substantially gained ground, whereas the value of community service
and social involvement declined. The present data for four Western
European countries now show that this feature may not have been
restricted to youths only, but was widespread across all cohorts.
If this holds, the shift should be interpreted as a period-effect
and not as a cohort-effect that effects youths only.