You are on page 1of 17

International Journal of Public Opinion Research Vol. No.

World Association for Public Opinion Research


; all rights reserved

THE TWO FACES OF


(POST)MATERIALISM:
A DECOMPOSITION APPROACH

Guy Moors

ABSTRACT
Ingleharts postmaterialism thesis always steered controversy. This paper focuses on the
internal consistency of the thesis by arguing that (post)materialism has two faces. It is
demonstrated that unfolding the Inglehart index into two separate indices leads to the
nding of dierential eects of education and cohort. Only the issues of giving people
more say versus maintaining order are autonomously related to cohort. Cohort dier-
ences as far as protecting freedom of speech and ghting rising prices are concerned
are virtually completely levelled of after controlling for education and marital status.
Education proved to be the single most important covariate in explaining these issues.

There are few studies on political culture that disregard Ingleharts thesis on
postmaterialism. Not that his thesis is generally accepted, on the contrary. The
core of his thesis has been repeatedly formulated in the course of his publications
(, , , , , and ) but can probably best be summarized in
essence in the dual-hypotheses model (). The rst hypothesis claims that
individuals priorities reect their socioeconomic environment: one attaches
relatively more importance to relatively scarce objects. This scarcity hypothesis
is supplemented by a socialization hypothesis that stresses the importance of
experiences in the so-called formative years: in reaching adulthood, values tend
to crystallize in personality. Taken together, these hypotheses form the clue to
the process of the Silent Revolution (Inglehart, ). The values climate (at the
aggregate or national level) changes gradually in time due to the process of social
metabolism (i.e. generation replacement): older generations die o and are
replaced by newer (younger) generations who hold dierent value orientations
because of historically dierent socioeconomic conditions in which they are raised.
In order to dene the type of values involved, Inglehart initially ( and )
referred to Maslows individual motivation theory () in which a distinction
is made between lower order needs of a physiological nature and higher social

I would like to express my sincere gratitude to the anonymous referees for their helpful comments. The article
was submitted to IJPOR September , . The nal version was received April , .
THE TWO FACES OF (POST-)MATERIALISM
or self-actualization needs. Values reecting lower order needs are labelled
materialism; social and self-actualization needs correspond with post-materialism.
It is important to stress that Ingleharts materialism may not be interpreted in
a rigid consumerist way, e.g. synonymous to having or property. Materialism
includes the notions of material and physical security. We also need to dierentiate
postmaterialism from anti-materialism. Postmaterialism reects values beyond
materialistic values rather than a refusal of them.

ISSUES OF CONTROVERSY: DEFINING THE RESEARCH


QUESTION
The internal logic of the Inglehart thesis has repeatedly been subject to controversy.
One of the rst issues that has been questioned is the assumption of relatively
stable values orientations both at the aggregate (cohort) and individual level.
Several authors incisively addressed the question (Dalton, ; Herz, ; van
Deth, ; Jagodzinski, ; Boeltken & Jagodzinski, ; de Graaf, ; de
Graaf, Hagenaars, & Luijkx, ; Klein, ), including Inglehart (, ,
, ). By now, the evidence looks impressive. At the cohort level, a drop
in postmaterialism in the s is followed by an equally signicant revival in
the s (Inglehart, ). In the early s postmaterialism slightly decreased
again (Inglehart & Abramson, ). These period uctuations appear to coincide
with the economic climate of the time. Furthermore, taking measurement error
and the ipsative property of his questionnaire into account, Inglehart is able to
report impressive correlations (ranging between . and .) between consecutive
measurements of postmaterialism from the same individuals (Inglehart, ). These
ndings are conrmed even by fairly critical researchers such as de Graaf ().
A second issue that arose scientic interest lies in the act of determining the
nature of the conditions that shape the socialization of values. One individual
characteristic, which has always been the subject of intense discussion, is the
positive association of higher levels of education with postmaterialism. Inglehart
interprets level of education in this regard as an indication of the relative
economic prosperity on a micro level: after all, the afuent are the ones who are
eligible for higher education. Several authors (Marsh, ; Laerty, ; van
Deth, ; and de Graaf, ) have quite rightly remarked that Inglehart is far
too eager in his neglect of the intrinsic characteristics of education: the higher the
level of education, the more one learns to stand up for oneself. Moreover, the
creation of a cultural capital equally has a distinct direct eect on the socialization
of postmaterialist values. Duch and Taylor () even claim that education is
the chief alternative explanation for the emerging values gap, rather than
cohorts. Their argument, however, is not that postmaterialism is learned in
educational institutions, but rather that the Inglehart index actually measures
pro-democratic orientations. As Warwick () indicates, the nature of the
INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH

relationship between education and postmaterialism is still subject to debate.


What Duch and Taylor wanted to demonstrate is that the validity of the post-
materialism scale can be questioned in reference to disputed causal eects.
Warwicks critical examination of the Inglehart thesis follows a similar perspective.
This brings us to the third issue of relevance to this paper. From the beginning,
there has always been some criticism about the way Inglehart operationalizes the
materialist/postmaterialist values dimension. What does it really measure? Is
one-dimensionality justied? And is the ranking technique appropriate? (Marsh,
; Herz, ; van Deth, ; Flanagan, ; Gehmacher, ; Hellevik,
; Bean & Papadakis, ; Brklin, Klein, & Russ, ; Sacchi, ; Davis
& Davenport, ; Davis, Dowley, & Silver, ).
The research question in this paper refers to the internal validity of the Inglehart
thesis. As such, we restrict our research to the key variables as they are delineated in
Ingleharts theory. However, we do not want to suggest that this approach is the
sole appropriate way of researching postmaterialism. A dierent, also valuable
and complementary approach would imply a kind of external validation of the
Inglehart thesis since it would extend the concept of postmaterialism to other
spheres of life and it would question the social denition of characteristics that
are assumed to bring about postmaterialism. In more recent work (), Inglehart
himself devoted more attention to this issue of external validation. What we want
to argue, however, is that even within the (perhaps narrow) context of the Inglehart
thesis, the internal consistency can be checked. Hence, we want to explore further
the disputed sources of preference for materialist versus postmaterialist items of
the Inglehart index (cf. Warwick, ) by balancing the relative importance of
education and cohort. We will demonstrate that Ingleharts postmaterialism
index can easily be decomposed into two separate indices that contrast a particular
materialist issue with a particular postmaterialist issue; and that cohort
membership and educational level are quite dierently related to each of these
two indices. As such we will argue that the generational argument is more heter-
ogeneous than Inglehart has suggested.

ANALYTICAL BASIS AND PROCEDURES


DATA
This paper draws on the same datasetthe Euro-Barometers1that Inglehart
repeatedly analyzed to develop his arguments. We have selected all respondents

1
Data are publicly available from ZUMA, Cologne, Germany, i.e. the European Communities Studies,
: cumulative le; ref. ICPSR /ZA ). Principal Investigators are R. Inglehart (University of
Michigan) and K. Reif & A. Mellich (European Commission). The data collection description indicates that
multiple probability sampling and stratied quota sampling are used, but does not provide information
regarding response rates.
THE TWO FACES OF (POST-)MATERIALISM
aged or older from nine countries (France, Belgium, the Netherlands,
Denmark, Luxembourg, (West) Germany, Italy, Great Britain, and Ireland) from
the pooled data (weighted N > ,). Intracountry weighting is used
to establish that data are representative at the national level. EC-weighting (as is
more frequently adopted by Inglehart), however, is not used. By consequence,
countries contribute to the pooled data equal to the number of respondents drawn
in the dierent samples. Except for Luxembourg, which always has a relatively small
sample size, the contribution of each country in the overall sample is about equal.
This dataset includes the original four-item question of the short index. The
wording is as follows:
There is a lot of talk these days about what the aims of this country should be for the next
ten years. On this card are listed some of the goals which dierent people would give top
priority. Would you please say which one of these you yourself consider the most important?
1. Maintaining order in the nation.
2. Giving people more say in important political decisions.
3. Fighting rising prices.
4. Protecting freedom of speech.
And which would be the next most important?
Respondents are classied as materialist if they combine options () maintaining
order and () ghting rising prices; and as postmaterialist if they choose ()
giving more say and () protecting freedom of speech. All other combinations
of preference dene the mixed intermediate category of the Inglehart index.
Education is measured as age when leaving school; the lowest level equals aged
or less (including no education); the highest level includes all ages higher or
equal to . A separate category is listed for respondents still attending school. In
this paper we recoded respondents who still attended school at the time of the
survey by imputing their age + year (assuming they would nish their education
within the year).2 Period (time of survey), birth cohorts, and country are the
other key covariates in the analyses. We also tested whether including other
covariates, such as income, occupational status, marital status, etc., explained
the relationship between cohorts and the values indices. It proved that only marital
status further reduced the impact of cohorts on values indices. As de Graaf
() has argued, marital status may be classied as a life-cycle characteristic.
For these reasons, marital status is included as an additional control variable.
Most of the eect of the individual economic indicators (income, occupation)
could be attributed to education, but cohort dierences remained virtually

2
Alternative coding (separate categories, assuming that every respondent aged or older would nish their
education after age ) did not substantially change the ndings reported in this paper. Leaving out these
respondents was not an option, since it implied retaining a highly selective sample (i.e. the lesser educated)
among the younger age groups.
INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH

unchanged when these characteristics were included. Given the fact that including
education, income, and occupation in one analysis may cause problems of multi-
collinearity, we decided not to include income and occupation. In most cases,
education is the causal antecedent, and hence the most obvious choice. To avoid
all misunderstanding, by choosing education, we are not suggesting that this
covariate merely indicates socioeconomic status. We refer to the aforementioned
discussion as plausible alternative interpretations of the concept of education.

PATTERNS OF CHOICE ON THE INGLEHART INDEX


As mentioned above, there has been quite some eort in previous research to
question the unidimensionality of postmaterialism. The discussion has a concep-
tual core, i.e. relative versus absolute values, that coincides with a methodological
dispute regarding the analysis of respectively ranking and rating questions.3 For
clarity sake, this paper is not yet another contribution to this discussion. In fact,
no factor analysis or other type of multidimensional scaling will be used. Moreover,
when we demonstrate the signicance of decomposing the Inglehart index into
two separate indices we do not claim to have proven the formers multidimen-
sionality. In fact, even adopting an appropriate scaling technique to the short
Inglehart questionnaire would merely conrm the unidimensionality. Instead
two other curiosities steered our research. The rst relates to the fact that Ingle-
hart is merely interested in two particular patterns of choice on his index, i.e. the
materialist pattern combining the issues ght rising prices and maintaining
order in the nation; and the postmaterialist choice of protecting freedom of
speech and giving people more say. Four other types of combination are all
classied as mixed types. Is there any justication for this? For this reason, this
section focuses on the pattern of choice on the four-item question in more detail.
Our second curiosity refers to the principal research question of this paper. If
we select the key covariates that Ingleharts theory delineates, can we expect that
their main eect is fully reected by their relationship to the Inglehart index? In
other words, are the eects of selected covariates similar for each issue involved in
the pattern of choice? These questions are answered in the next section.
Researching the relationship between the rst and second choice on the four-item
question is not straightforward given the fact that we are dealing with ipsative
measurement. After all, a classic -way frequency table of rst by second choice
would include structural zeros on the main diagonal since any rst item chosen is
excluded as an option in the second preference question. Hence, structural zeros
have to be imposed if we want to research the pattern of association between rst

3
Also, in his later work (), Inglehart has picked up this discussion favoring, of course, approaches that
conrm his thesis. Purely from a methodological point of view and focusing on the set of questions he has
developed, evidence does seem to conrm his arguments. If one takes the characteristic measurement errors of
each type of questions into account, rating and ranking techniques produce similar results.
THE TWO FACES OF (POST-)MATERIALISM
TABLE Standardized residuals (observedexpected frequencies) for particular
combinations of rst and second preference on Ingleharts four-item index
Panel A: Residuals of six choice patterns
Inglehart index Choice patterns Standardized residuals
Postmaterialism Say + Freedom .
Say + Prices .
Mixed Prices + Freedom .
Order + Say .
Materialism Order + Freedom .
Order + Prices .
Panel B: Frequency table of residuals
First choice
Order Say Prices Freedom
Second choice Order * . . .
Say . * . .
Prices . . * .
Freedom . . . *
Note: Structural zeros imposed on main diagonal.

and second choice. In Table we present the standardized residuals calculated from
the dierences between the observed and expected frequencies after dening the
main diagonal as structural zeros. Expected frequencies are calculated from the
independence model assuming no association between rst and second choice, given
the structural zeros. These standardized residuals can be calculated for any combi-
nation (see frequency table of residuals) but we summed the residuals for particular
combinations ignoring which of the two items was chosen as rst or second priority.
The rst observation conrms Ingleharts arguments; i.e. the largest positive
residuals are in consistency with the materialist/postmaterialist poles of his index.
As such, this analysis reafrms previous ndings regarding the association between
item preferences. The second nding, however, that leaps to the eye is that the
largest negative residuals are consistently found among two of the four patterns of
choices that are classied as mixed types in Ingleharts index. The combinations
ghting rising prices with freedom of speech, and maintaining order with giv-
ing people more say are less frequently observed than could be expected if statisti-
cal independence would have been observed.4 This indicates that within each of

4
Note that Flanagan () also noticed a conceptual dierence between the two materialist issues. However
sound his theoretical arguments are, his analysis is awed because it fails to recognize response set bias. Herz
() used multidimensional scaling and also found that freedom of speech and ghting rising prices were
at the greatest distance from one another. However, the distance from the former issue to the other materialist
issue of maintaining order is also pronounced.
INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH

these pairs the items function as each others opposite. That maintaining order
and giving people more say are dissonant choices does not come as a surprise
given Duch and Taylors () remark that the Inglehart index reects demo-
cratic orientations. As far as ghting rising prices versus freedom of speech is
concerned, we can only speculate. To some extent, of course, this may be merely
an instrumental outcome of the fact that maintaining order and giving people
more say are logically dissonant choices, i.e. the remaining two items become
dissonant by consequence of question wording. On the other hand, the issue of
ghting rising prices implies a certain level of government interference, which
may be associated with the center left. We found supportive evidence for the latter
argument in the fact that voting for Socialist parties was positively associated with
ghting rising prices in France, Belgium, the Netherlands, Italy, Luxembourg,
and Great Britain. The Social Democratic or Christian Democratic party vote in
Belgium, Germany, Italy and Denmark was also related to greater preference for
ghting rising prices. The short index, however, has no complementary center-
right issue, but semantically freedom probably comes closest. Nevertheless,
regardless of any internal logic, this pattern of two sets of dissonant items has been
observed consistently for all countries and every period.5

DECOMP OSING T H E P O S T M A T E R I A L I S M I N D E X
As is commonly known, Inglehart denes his short index as the percentage of post-
materialists minus the percentage of materialists. Many people still think of this index
as an aggregate level measure that needs to be calculated from estimating each of these
percentages separately.6 In fact, the index has been criticized (see Inglehart, )
because it left out the large mixed category that could distort the analysis of cohort
dierences in time. However, they are mistaken, since any score on this Inglehart
index can be calculated as the mean score on the individual level index. This index
assigns the scores of , and to respectively a materialist, mixed, and
postmaterialist combination of rst and second choices on the four-issue question.
A second feature of the Inglehart index, of whichto the best of our know-
ledgemany researchers are unaware, is the possibility to decompose this Inglehart
index into two indices (Moors, , ) contrasting a particular materialist item
with a postmaterialist item (see Table ). Taken together they exactly dene the
Inglehart index. Consistent with the ndings presented in Table , these two
indices contrast respectively freedom of speech with ghting rising prices and
more say with maintaining order. For each of these two indices we assigned
a score of if the materialist item was chosen (rst or second choice) and + if
the postmaterialist item was chosen. Again a mean score on these indices can be

5
Analyses
not reported, but available by request from the author.
6
Even Inglehart () is ignorant of the fact that this is not necessary.
THE TWO FACES OF (POST-)MATERIALISM
TABLE A decomposition of the Inglehart index. Illustrative example of
crosstabulation of the rst by the second choice on Iongleharts four-item index
First choice
Order Say Prices Freedom Total
Second choice Order
. . . .
Say
. . . .
Prices
. . . .
Freedom
. 8.6 . .
Total N
Percent . . . . .
Note: Entries are counts, cell percentages in italics.
Calculations:
() Inglehart index: Percent postmaterialist percent materialists
(. + .) (. + .) = .
() Index percent freedom of speech percent ghting rising prices
(. + .) (. + .) = .
() Index percent give people more say percent maintain order
( + .) (. + .) = .
() Inglehart index as mean of index () and ()
{[(. + .) (. + .)] + [( + .) (. + .)]}/ = .

interpreted as the percentage choosing a particular postmaterialist item minus the


percentage choosing the dissonant materialist item. Taking the mean of the sum of
these two decomposition indices reproduces exactly the score on the Inglehart index.
In what follows we will demonstrate that the two indices reect two dierent
faces of (post)materialism. One should keep in mind, however, that our approach
is fundamentally dierent from those researchers who argued for a multidimen-
sional approach. Given the properties of the Inglehart index and the two separate
indices explained in this section, we can restrict research to analyses with each
item as dependent variable (see below) and just fold up these results (see Figure )
in order to reproduce exactly the same results Inglehart would have had if only his
index were used as the dependent variable. To the best of our knowledge, no other
approach allows for such an exact reproduction of the short Inglehart index.

RE-EXAMINING THE COHORTEDUCATION


STRATIFICATION OF POSTMATERIALISM
The key issue in this section is the question of to what extent the heart of Ingleharts
thesis regarding the eect of cohorts on values is reconrmed in the case of the
INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH

two separate indices. We do not pretend to present models that are complete in
the sense that they include all possible explanations of postmaterialism. Such
models should include both aggregate and individual level characteristics since
postmaterialism is linked to both personal as well as contextual historical circum-
stances. As mentioned in the introduction, we limit our research to the question
of the relative importance of education compared to the cohort stratication of
values. As will be demonstrated, even such simple models, including only a few
covariates, reveal the particularity of the cohort dierences and the signicance
of education.
But can we expect cohort dierences to dier depending on the two separate
scales? Evidently, Inglehart has not explicitly raised this question, but re-examining
the original thesis could lead to some suggestions. We discern an evolution in
Ingleharts work that could lead to some tentative hypotheses. As mentioned in
the introduction, the original theoretical frame of reference (Inglehart, ,
) was based on Maslows individual motivation theory, in which lower
physiological needs (physical and material security) are contrasted with higher
social and self-actualization needs. On the basis of the scarcity hypothesis, two
historic conditions were attributed equal importance in the rst publications: the
increasing economic prosperity and the absence of direct war experience. Later
on Inglehart attached increasing importance to the economic conditions (cf. van
Deth, ). By , the Maslowian needs theory is merely complementary to
the principle of marginal utility in economic theories. The impact of the
economic conditions is given extensive attention in this re-examining of the theory
of value change (Inglehart, ), whereas the direct war experience is no longer
mentioned and the concept of physical security features in the discussion in a
rather inconspicuous way. Nevertheless, the Inglehart thesis seems to imply that
economic conditions primarily shape the cohort stratication of values, whereas
the absence of war distinguishes pre- from post-war born cohorts (cf. de Graaf,
). The two materialist items of the short index explicitly refer to the afore-
mentioned historical conditions. Economic or material security is reected in the
issue of ghting rising prices, whereas maintaining order is referring to physical
security.
Given these arguments, there is reason to believe that the cohort stratication
with regard to the economic aspect will be more pronounced than the stratication
for the non-economic aspect. The reason is that the economic security aspect is
inuenced by both socio-historical conditions, whereas physical security (measured
by the need for social order) basically relates to the absence of war experience.
Hence a tentative hypothesis could be that the cohort dierences may be larger
in the case of the freedom versus prices index compared to the say versus
order index. To the extend that Inglehart was right in interpreting educational
levels as measuring parental afuence, we may expect that cohort dierences in
freedom versus prices will decrease if education is taken into account. Education
THE TWO FACES OF (POST-)MATERIALISM
itself should be fairly important. Following Duch and Taylors arguments
education should also be an important factor in explaining dierences in say versus
order.

RESULTS: THE TWO FACES OF POSTMATERIALISM


In the following sections we summarize the results of a series of dummy regression
analyses7 with the four items (maintaining order, more say, rising prices, and
freedom of speech)8 and the two indices (freedom of speechghting rising
prices; and more saymaintaining order) as dependent variables into two
gures. Each of these gures represents mean scores on these items and indices
for dierent cohorts, periods (year of survey) and educational levels. We compare
observed values (unadjusted for covariates) with two models that include control
variables (adjusted for covariates). The rst multivariate model includes the
aforementioned covariates and dummies per country, whereas in the second
model marital status is added as an additional covariate. All values presented are
predicted values (percentages or dierences in percentage).

T H E F I R S T F A C E : C O N T R A S T I NG F R E E D O M O F
SPEECH WITH FIGHTING RISING PRICES
Contrary to our expectations the cohort dierences on the index freedom vs. prices
almost completely vanish after controlling for education and marital status
(Figure ).9 Cohort dierences range within a percent margin of dierence.
This nding is reconrmed for each of the two constituting items, i.e. the largest
dierence on either issue is less than percent. Especially in the case of ghting ris-
ing prices, the autonomous eect of cohorts is remarkably low given the observed
dierences. Note that Flanagan () was one of the rst to detect that materialism
in the strict economic sense of the word is not related to age. His analysis, however,
does not include the ranking question Inglehart uses to measure materialism.
Contrary to the small ignorable eect of cohorts, educational and period
eects are pronounced and fairly robust. The longer one attends school, the
more likely one prefers freedom of speech relative to ghting rising prices. We
could speculate on what causes education to be the prime stratier of this values
index, and it is most likely that it is a multitude of distinct factors that is tapped

7
Since all covariates were treated as nominal, we applied multiple classication analysis that is more exible
in handling nominal covariates and is, at the same time, similar to dummy regression, except that it uses
deviation coding rather than dummy coding. Note that, since we treat all variables as nominal (i.e. sets of
dummy variables), the parameter estimates all fall within the admissible range (taken on values between and
and indicating percent).
8
An item was coded if it was chosen as a rst or second option; a value of was assigned when the
item was not chosen. Hence, a mean score on an item can also be interpreted as percent.
9
The impact of country and period proved to be only marginal (analysis not reported).
INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH

FIGURE Cohort, education, and period dierences in freedom of speech and rising
prices

by level of education. What cannot be ignored though, is the fact that education
almost single-handedly explained cohort dierences, leaving little room to
conclude that intergenerational dierences are robust. Moreover, the autono-
mous trend eect is far more pronounced, suggesting that this part of the original
postmaterialism index is highly sensitive to temporal eects.
Inglehart (, Inglehart & Abramson, ) repeatedly referred to ination
as the primary cause of period uctuations, i.e. as the ination rate rises, the
postmaterialism index drops and vice versa. According to Inglehart the rationale
for expecting this relationship is the relatively obvious linkage between high levels
of ination and the goal of ghting rising prices (Inglehart & Abramson, ,
p. ). The trend eect reported in Figure seems to conrm this explanation,
although not perfectly. Trend eects are most pronounced as far as ghting rising
prices is concerned. However, the aforementioned rationale that links this trend
to levels of ination is less consistent than Inglehart has reported in reference to
trends in postmaterialism. In the period before as well as after ination
rates increased, whereas the preference for ghting rising prices was diminishing
(see Figure in the Appendix). In Figure we will demonstrate that the
relationship of period eects of Ingleharts materialismpostmaterialism index
and trends in ination has to do with the fact that in the pre- and post-
era the preference for the other materialist issue of maintaining order increased.
Hence, the relative share of ghting rising prices and maintaining order in
dening the materialist category of the Inglehart index altered in these two periods.
Finally, it is worth mentioning that some country specic patterns10 emerged
regarding how the post-war cohorts diered on the freedom vs. prices index.

10
Figures not reported, results can be made available by the author, on request.
THE TWO FACES OF (POST-)MATERIALISM
In the Netherlands, Denmark, and (to a lesser degree) France, the index sharply
decreased, suggesting that cohorts became less post-materialist as far as this part
of the Inglehart index is concerned. In Germany, Italy, and (to a lesser degree)
Ireland a reversed tendency emerges: post-war cohorts became more postmaterialist.
As such, the ndings from these countries are more consistent with Ingleharts thesis.

THE SECOND FACE: CONTRASTING GIVING PEOPLE


MORE SAY WITH MAINTAINING ORDER IN THE
NATION
Ingleharts postmaterialism thesis appears to be nearly perfectly reected in the
pattern of associations with the second index more say vs. maintaining order
(Figure ). Cohort dierences are clearly observed, andeven more importantly
hardly diminish after controlling for other covariates. The cohort stratication
on the issue of maintaining order, in particular, remains virtually unaected by
including additional covariates. However, these cohort dierences should be
evaluated with period eects in mind. We briey commented on this issue in the
previous section by drawing attention to the nding that the preference for
maintaining order increased in the period before and after . In the
period in between, the preference remained fairly stable. A researcher focusing
only on the materialist issue of maintaining order could make a point by arguing
that the cohortperiod prole indicates life-cycle eects rather than generational
eects. After all, preference regarding maintaining order decreases at the
cohort level, and at the same time increases by period, indicating that as cohorts
mature they become more conservative on the issue. The evidence from the second
postmaterialist issue of giving people more say, however, does not support such
a conclusion. In this case the preference increases with both cohort and period,
which is consistent with Ingleharts expectations.
One question emerges from the data: is a reversal in orientation happening
among the younger cohorts? The two youngest cohorts born in the s are
more conservative, i.e. combining a higher preference for maintaining order
with a lower preference for giving more say than the cohort, which has
the highest postmaterialist preference on the more say vs. maintaining order
index of all cohorts. If this pattern continues to show up, Ingleharts silent revo-
lution may be reversed, and the rise in postmaterialism due to generational
replacement will drop in the long run. In running separate analyses per country,11
we found that countries may exhibit dierent patterns in the future. Diering
from the overall pattern, the youngest cohorts in Germany, Ireland, and Great
Britain held the highest scores on the more say vs. maintaining order index.

11
Figures not reported, results can be made available by the author, on request.
INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH

FIGURE Cohort, education, and period dierences in more say and maintaining
order

Hence we cannot (yet) conclude that a reversal in orientation is consistently


emerging in Europe.
Education is related to the more say vs. maintaining order index in the way
we expected, i.e. preference for more say increases relative to the preference for
maintaining order as education rises. Dierences between the categories that
nished education by the age of , however, vanish after controlling for other
covariates. The category that really stands out is the highest level of education.
Hence, it is safe to conclude that postmaterialism is clearly linked to the highest
level of educational status as far as the second say vs. order index is concerned.

FOLDING UP
In Figure we bring together the divergent eects of education and cohort in
a two-dimensional chart, to illustrate how these issue-specic eects sum up to
the original Inglehart index. The X-axis is dened by the index percent freedom
of speechpercent ghting rising prices; the index percent more saypercent
maintaining order denes the Y-axis. The main diagonal of this -dimensional
plot denes the third Z-axis that indicates the coordinates on the Inglehart
index. After all, we have demonstrated that the Inglehart index is equal to the
mean score on the sum of the two composite indices. In Figure , for instance, the
(x;y) coordinates of the cohort born between and equals (.;.).
The corresponding value on the Inglehart index equals . [=(. .)/].
Figure is, of course, a summary of the analyses presented in the previous
sections. Connecting cohorts produces nearly vertical lines, indicating that
cohort is virtually exclusively related to the index more saymaintaining order
THE TWO FACES OF (POST-)MATERIALISM

FIGURE Two-dimensional plot of the eects of cohort and education on the Inglehart
index

(Y-axis) and only slightly related to the index freedom of speechghting rising
prices (X-axis). Connecting educational levels reveals lines with an angle of
approximately degrees. This indicates that education is related to the two
composite indices, but more strongly in the case of the X-axis (freedom of
speechghting rising prices). The dierence between the highest and lowest
level of education is nearly twice as much on this index than on the index more
saymaintaining order.
INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH

CONCLUSION
The fact that two particular combinations of rst and second choice on the short
postmaterialism questionnaire are less favored is not a coincidence. In fact this
nding is as signicant as demonstrating that respondents tend to choose either
a materialist or a postmaterialist preference. This paper has demonstrated that
(post-)materialism has two faces and that the Inglehart index can easily be
decomposed into two separate composite scales, i.e. one index contrasting the
issues more say with maintaining order and a second index contrasting freedom
of speech with ghting rising prices. The pattern of association between rst
and second choice on the short postmaterialism questionnaire has validated this
decomposition. Furthermore, we found that cohort and education are dierently
related to these two composite indices. Only as far as the more saymaintaining
order index is concerned did cohorts exhibit an autonomous eect. Cohort
dierences on the other index freedom of speechghting rising prices nearly
vanished after controlling for a small set of relevant covariates. Education, on the
contrary, proved to be the single most important covariate of this latter index.

AP P EN DI X

FIGURE Trend in ghting rising prices compared to ination rate

REFERENCES
Bean, C., & Papadakis, E. (). Polarized priorities or exible alternativesDimen-
sionality in Ingleharts materialism-postmaterialism Scale. International Journal of
Public Opinion Research, , .
THE TWO FACES OF (POST-)MATERIALISM
Boeltken, F., & Jagodzinski, W. (). Postmaterialism in the European community,
: Insecure value orientations in an environment of insecurity. Comparative
Political Studies, , .
Brklin, W., Klein, M., & Russ, A. (). Dimensions of value changeAn empirical
investigation of the dimensions and the dynamics of change of social value orientations.
Politische Vierteljahresschrift, , .
Dalton, R. J. (). Was there a revolution? A note in generational versus
life-cycle explanations of value di erences. Comparative Political Studies, ,
.
Davis, D. W., & Davenport, C. (). Assessing the validity of the postmaterialism
index. American Political Science Review, , .
Davis, D. W., Dowley, K. M., & Silver, B. D. (). Postmaterialism in world societies:
Is it really a value dimension? American Journal of Political Science, , .
Deth, J. W. van. (). Politieke waarden : een onderzoek naar politieke waardenorintaties
in Nederland in de periode tot en met . Amsterdam: CT-Press.
Duch, R. M., & Taylor, M. A. (). Postmaterialism and the economic condition.
American Journal of Political Science, , .
Flanagan, S. C. (). Value change in industrial societies. American Political Science
Review, , .
Gehmacher, E. (). Was misst die Inglehart-Skala fur Postmaterialismus? sterreichische
Zeitschrift fr Soziologie, , .
Graaf, N. D. de (). Postmaterialism and the stratication process. An international
comparison. Utrecht: ISOR.
Graaf, N. D. de, Hagenaars, J., & Luijkx, R. (). Intragenerational stability of
postmaterialism in Germany, the Netherlands and the United States. European
Sociological Review, , .
Hellevik, O. (). Postmaterialism as a dimension of cultural change. International
Journal of Public Opinion Research, , .
Herz, T. A. (). Der Wandel von Wertvorstellungen in westlichen Industriegesells-
chaften. Klner Zeitschrift fr Soziologie und Sozialpsychologie, , .
Inglehart, R. (). The silent revolution in Europe: intergenerational change in
post-industrial societies. American Political Science Review, , .
Inglehart, R. (). The silent revolution. Princeton, NJ: Princeton University Press.
Inglehart, R. (). Post-materialism in an environment of insecurity. American Political
Science Review, , .
Inglehart, R. (). Aggregate stability and individual level ux in mass belief systems:
the level of analysis paradox. American Political Science Review, , .
Inglehart, R. (). Culture shift in advanced industrial society. Princeton, NJ: Princeton
University Press.
Inglehart, R. (). Modernization and postmodernization. Cultural, economic, and political
change in countries. Princeton, NJ: Princeton University Press.
Inglehart, R., & Abramson, P. R. (). Economic security and value change. American
Political Science Review, , .
Jagodzinski, W. (). Wie transformiert man labile in stabile Relationen? Zur Persistenz
postmaterialistischer Wertorientierungen. Zeitschrift fr Soziologie, , .
INTERNATIONAL JOURNAL OF PUBLIC OPINION RESEARCH

Klein, T. (). Zur Bedeutung von Alters-, Perioden- und Generationseinussen


fr den Wandel politischer Werte in der Bundesrepublik. Zeitschrift fr Soziologie, ,
.
Laerty, W. M. (). Basic needs and political values. Some perspectives from Norway
on Europes silent revolution. Acta Sociologica, , .
Marsh, A. (). The silent revolution, Value priorities and the quality of life in Britain.
American Political Science Review, , .
Maslow, A. (). Motivation and personality. New York: Harper & Row.
Moors, G. (). Generational dierences in political value orientations: in interna-
tional comparison. In H. A. Becker (Ed.), Dynamics of cohort and generations research.
Proceedings of a Symposium held on , , and December at the University
of Utrecht, The Netherlands (pp. ). Amsterdam: Thesis Publishers.
Moors, G. (). The dynamics of values-based selection and values adaptation. With an
application to the process of family formation. Ph.D. Dissertation, IPD-Working Paper
, Interface Demography, VUB.
Sacchi, S. (). The dimensionality of postmaterialism: an application of factor analysis
to ranked preference data. European Sociological Review, , .
Warwick, P. V. (). Disputed cause, disputed eect: the postmaterialist thesis
re-examined. Public Opinion Quarterly, , .

BIOGRAPHICAL NOTE
Guy Moors holds a Ph.D. in sociology and is an associate professor (U.D.) at the
Department of Methodology & Statistics at Tilburg University in the Netherlands. His
research interests include values research, measurement of attitudes, social demography,
ethnic minorities, and longitudinal data analysis.
Address correspondence to Guy Moors, MTO-FSW room S, Tilburg University,
P.O. Box , LE TILBURG, The Netherlands, E-mail: guy.moors@uvt.nl

You might also like