The current issue and full text archive of this journal is available on Emerald Insight at:

www.emeraldinsight.com/2059-5794.htm

What people
What would people do with their would do with
money if they were rich? their money if
they were rich
A search for Hofstede dimensions
across 52 countries 93

Michael Minkov Received 29 November 2018

Revised 20 February 2019
Sofia Campus, 1 March 2019
Varna University of Management – Sofia Local Center, Sofia, Bulgaria and Accepted 6 March 2019

University of Tartu, Tartu, Estonia

Pinaki Dutt
Department of Strategy, MediaCom, Singapore
Michael Schachner
Itim International, Helsinki, Finland
Janar Jandosova
Sange Inc., Astana, Kazakhstan
Yerlan Khassenbekov
AIRI, Astana, Kazakhstan
Oswaldo Morales
ESAN University, Lima, Peru, and
Vesselin Blagoev
Varna University of Management – Sofia Local Center, Sofia, Bulgaria

Abstract
Purpose – The purpose of this paper is to test the replicability of Hofstede’s value-based dimensions –
masculinity–femininity (MAS–FEM) and individualism–collectivism (IDV–COLL) – in the field of consumer
behavior, and to compare cultural prioritizations with respect to disposable income budgets across the world.
Design/methodology/approach – The authors asked 51,529 probabilistically selected respondents in 52
countries (50 nationally representative consumer panels and community samples from another two countries)
what they would do with their money if they were rich. The questionnaire items targeted Hofstede’s MAS–
FEM and IDV–COLL as well as a wider range of options deemed sufficiently meaningful, ethical and moral
across the world.
Findings – The authors obtained two main dimensions. The first contrasts self-enhancing and altruistic
choices (status and power-seeking spending vs donating for healthcare) and is conceptually similar to
MAS–FEM. However, it is statistically related to Hofstede’s fifth dimension, or monumentalism–flexibility
(MON–FLX), not to MAS–FEM. The second dimension contrasts conservative-collectivist choices and
modern-hedonistic concerns (donating for religion and sports vs preserving nature and travel abroad for
pleasure) and is a variant of COLL–IDV.
Research limitations/implications – The authors left out various potential consumer choices as they were
deemed culturally incomparable or unacceptable in some societies. Nevertheless, the findings paint a sufficiently
rich image of worldwide value differences underpinning idealized consumer behavior prioritizations.
Practical implications – The study could be useful to international marketing and consumer behavior experts.
Social implications – The study contributes to the understanding of modern cultural differences across
the world.
Originality/value – This is the first large cross-cultural study that reveals differences in values through Cross Cultural & Strategic
a novel approach: prioritizations of consumer choices. It enriches the understanding of IDV–COLL and Management
Vol. 26 No. 1, 2019
MON–FLX, and, in particular, of the value prioritizations of the East Asian nations. The study provides pp. 93-116
© Emerald Publishing Limited
2059-5794
The data collection for this study was financially supported by MediaCom Ltd. DOI 10.1108/CCSM-11-2018-0193
CCSM new evidence that Hofstede’s MAS–FEM is a peculiarity of his IBM database with no societal analogue.
Some of the so-called MAS–FEM values are components of MON–FLX, which is statistically unrelated
26,1 to Hofstede’s MAS–FEM.
Keywords Consumer behaviour, Hofstede’s cultural dimensions, Collectivism/Individualism,
Masculinity/Femininity, Monumentalism/Flexibility
Paper type Research paper

94 Introduction
Approaches to the study of values in the cross-cultural literature
Large-scale studies of national culture have often focused on values, although their
approaches have been different. Schwartz (1994, 2008a) defined values as trans-situational
goals and operationalized them in terms of items that ask respondents what is important in
their own lives. Hofstede (2001) used a more situational approach: he focused on personal
work-related goals in a specific multinational company, believing that these goals would
reflect societal values. Project GLOBE (House et al., 2004) operationalized values in an
unusual way: not as personal goals but as ideologies about desirable behaviors or
personality traits that all people in society should ideally share. Two other approaches are
worth mentioning, too. Minkov, Bond, Dutt, Schachner, Morales, Sanchez et al. (2018) and
Minkov et al. (2017) analyzed national cultures through self-construals. Many of these
obviously reflect personal values. For instance, respondents who select a statement such as
“I like to help people” evidently value helping, whereas those who choose “I am a very
religious person” certainly value religion. There is yet another approach to the study of
national values, pioneered by the World Values Survey, subsequently developed by Minkov,
Dutt, Schachner, Jandosova, Khassenbekov, Morales et al. (2018), asking the respondents to
describe the values and desirable traits that they wish to inculcate in their children.

A novel approach: studying values through the behavioral inclinations of consumers

Recently, Tung and Stahl (2018) encouraged researchers to use new designs in studies of
culture and pay more attention to situational factors. This paper is partly a response to that
call. While we still use the old method of self-reports, the novelty in our study is its object.
We ask large panels of probabilistically selected consumers in 50 countries, and large
community samples from another two countries, what they would do with their money if
they had so much of it that they felt rich. This is a novel way of studying values.
The statements that the respondents provided in our study can be called behavioral
inclinations. These inclinations are motivated by values since the latter are guiding
principles that drive behavior. According to one of the early definitions of values, “a value is
a conception, explicit or implicit, distinctive of an individual or characteristic of a group, of
the desirable which influences the selection from available modes, means, and ends of
action” (Kluckhohn, 1967, p. 395). And according to Schwartz and Bardi (2001), values serve
as guiding principles in people’s lives. Thus, values may drive behavior. However,
situational factors may override values. Swidler (1986) pointed out that in a “culture of
poverty” one may take a course of action that is not consistent with the typical actions of the
affluent classes even if poor and well-to-do people hold similar values. This is especially true
with respect to consumer behavior. Different social classes may value a particular expensive
product, such as a French perfume or a luxury car, to the same extent because it conforms to
their personal values but the needy are unable to afford such products. To place all
respondents in a similar, albeit hypothetical situation, we asked them all to imagine that
they were rich. One may argue that people who have never experienced wealth may not
have a very good idea of how they would spend their money if they became rich. This issue
cannot be resolved without an empirical test. If our study yields uninterpretable results, one
of the explanations may be that many respondents made haphazard statements. However, if
we obtain patterns that resemble previously reported dimensions of culture, our conclusion What people
must be that we have measured some real cultural substance. would do with
Our study could make several contributions. First, it could be useful to researchers in the their money if
field of cultural differences in consumer behavior. The association between national cultural
differences and national differences in consumer behavior has been studied extensively. they were rich
Dimensions of national culture have been found to predict or mediate differences in
cross-national product diffusion rates (Dwyer et al., 2005), acceptance of new products in a 95
country (Yeniyurt and Townsend, 2003), intention to purchase personalized products
(Moon et al., 2008), pro-environmental consumer behavior (Soyez, 2012), e-commerce
adoption (Pavlou and Chai, 2002), adoption and diffusion of internet retailing (Gong, 2009),
customer satisfaction in the hospitality industry (Radojevic et al., 2017), consumer
evaluations of travel services (Crotts and Erdman, 2000), impulsive buying behavior (Kacen
and Lee, 2002) and consumer complaint behavior (Liu and McClure, 2001) among others. It
has also been argued that the cultural differences accounting for differences in consumer
behavior are stable and unlikely to disappear even if national income levels converged
(De Mooij and Hofstede, 2002).
The typical approach in studies of national culture and consumer behavior is to seek
correlations between value-based indices of national culture, such as those of Hofstede (2001)
or Schwartz (2008b), and measures of consumer behavior. However, Hofstede’s indices are
actually measures of work goals in a specific multinational company, whereas Schwartz’s
indices reflect broad personal and societal values. There are no measures of national culture,
across a wide range of countries, extracted directly from values or value-related statements in
the context of consumer behavior. The goal of our study is to fill this gap.
If our search for such dimensions is successful, the results can have important practical
consequences for practitioners in the field of marketing because it will equip them with
better knowledge of the way that cultural differences account for consumer behavior
differences. This should result in more effective marketing strategies, taking into account
local cultural factors. A discovery of dimensions of national culture that are direct
reflections of differences in values related to consumer behavior would also have important
theoretical implications. It would provide the first unequivocal confirmation of the
hypothesis that cultural values are associated with consumer behavior. More than 40 years
ago, Henry (1976) noted that although marketing theorists generally accept that culture is
one of the determinants of consumer behavior, this relationship is based on intuitive
assumptions and is not supported by empirical evidence. Although such evidence exists by
now, the studies that provide it were typically executed in a small number of countries, or
focus on the relationship between existing dimensions of culture and a very specific
consumer behavior. There is no large cross-cultural study that elucidates the relationship
between broad cultural values and broad patterns of consumer behavior. However, we must
again stress the need to heed Swidler’s (1986) warning. Values drive behavior – and this is
obviously true of consumer behavior, too – only in the absence of restrictive circumstances,
such as poverty, that make it impossible for people to act in accordance with their values.
Therefore, how closely our measures of values will be correlated with actual consumer
behavior statistics will depend on situational factors.
This paper could also make a theoretical contribution concerning Hofstede’s model of
national culture. In his criticism of that model, Minkov (2018) argued that it should be
reduced to two dimensions of culture: individualism–collectivism (IDV–COLL) and
monumentalism–flexibility (MON–FLX), which is a reconceptualization of Confucian
dynamism (Chinese Culture Connection, 1987) or long-term orientation (LTO) (Hofstede,
2001). These dimensions have been replicated in studies by different authors, most recently
by Minkov et al. (2017) and Minkov, Bond, Dutt, Schachner, Morales, Sanchez et al. (2018),
whereas a search for masculinity–femininity (MAS–FEM) and uncertainty avoidance
CCSM replications proved fruitless (Minkov, 2018). Subsequently Minkov, Dutt, Schachner,
26,1 Jandosova, Khassenbekov, Morales et al. (2018) replicated this two-dimensional model in an
analysis of parental advice for the socialization of children. It is interesting to establish
whether a similar model would emerge in a study of behavioral inclinations in the field of
consumer behavior.
Alternatively, it is possible that, despite the criticisms, Hofstede’s most controversial
96 value dimension – MAS–FEM – will emerge in this study of values related to consumer
behavior. In his main academic book, “Culture’s Consequences,” Hofstede (2001) devoted a
whole sub-chapter to associations between MAS–FEM and consumer behavior, citing
analyses by Dutch consumer behavior consultant Marieke de Mooij. According to that
chapter, status purchases are more typical in MAS cultures, where people are more likely
than those in FEM ones to own expensive watches and jewelry, and to fly in business class
for pleasure. It is important to note that these associations are actually zero-order
correlations across European countries only, which have neither been shown across a more
representative sample of the world’s cultures, nor tested in regression models together with
other potential predictors. We do not intend to do either of the two in this study. However, as
a first step, we will seek an answer to a more basic question: Does a dimension of national
culture similar to MAS–FEM emerge from an analysis of consumers’ inclinations to spend
their disposable income in various ways, including inclinations to buy status products?

A brief overview of MAS–FEM theory and the analysis through which it was created
In this section, we briefly explain Hofstede’s MAS–FEM theory as it provided a foundation
for our study. That theory was derived from an empirical study of work goals across the
national subsidiaries of the IBM company around 1970. Subsequently, it was extrapolated to
the level of whole societies, based on Hofstede’s assumption that the IBM differences mirror
national differences.
Hofstede factor-analyzed 14 work goals in the IBM survey. It is important to note that
before the analysis the scores were ipsatized: a procedure that transforms raw scores into
prioritizations. In our study, we used a procedure that amounts to ipsatization and thus
preserved the spirit in which MAS–FEM was created. Without ipsatization, it may be
impossible to replicate MAS–FEM because ipsatization may produce effects that are not
observed when raw scores are used. In particular, ipsatization may transform positively
correlated items into negatively correlated opposites. For example, without ipsatization
MAS and FEM goals and values are not opposites.
Hofstede’s analysis produced an IDV–COLL factor and a MAS–FEM factor. The
simultaneous emergence of these two dimensions in the same factor analysis is also
noteworthy, especially in view of the fact that some items cross-loaded. In other words, they
were associated with both dimensions. When this is the case, it may be impossible to get a
very close replication of either factor on its own if one factor-analyzed only the items with
high loadings on a particular factor. In other words, IDV–COLL and MAS–FEM are like
conjoint twins that may not be able to exist separately in the exact form in which they were
created. In fact, there exist replications of IDV–COLL that did not include MAS–FEM items
in the analysis (although none of them are very close to Hofstede’s IDV–COLL version) but
there are no replications of MAS–FEM at all. It is possible to speculate that one reason
for this failure may be the fact that Hofstede’s procedure, combining IDV–COLL and
MAS–FEM items in the same factor analysis, has never been followed. The findings of our
study, however, demonstrate that we did follow the spirit of Hofstede’s procedure.
The MAS–FEM factor in Hofstede’s analysis consisted mainly of importance of earnings,
recognition and advancement (MAS goals) vs importance of good relationships with one’s
manager and cooperation with one’s colleagues (FEM goals). Of note, Hofstede (2001) saw the
MAS pole of the dimension as associated with self-enhancement, because earnings, recognition
and advancement boost one’s ego. Extrapolating his IBM work-goal findings to the level of What people
whole nations, he saw MAS societies as oriented toward competition and achievement. would do with
The latter needs to be advertised, hence a desire to display status symbols in MAS cultures. their money if
FEM societies downplay the importance of competition and achievement, and consequently of
status symbols. Instead, they value cooperation and good human relationships. they were rich

Criticisms of MAS–FEM 97
In his public appearances around the world, Hofstede has often spoken about the backlash
that the MAS and FEM labels have produced, especially in the English-speaking countries.
According to him, his terminology has been accused of reinforcing sexist stereotypes.
A terminology can easily be replaced if it offends somebody but an important question
would remain. Can Hofstede’s findings from his IBM study really be taken as representative
of whole societies? Minkov (2018) provided evidence from a number of studies and literature
reviews demonstrating quite clearly that the IBM database did not provide a close
representation of national cultural differences and was even very misleading in some cases.
For instance, male–female differences (or distances) in values were associated with
MAS–FEM differences in the IBM database: in more MAS countries, men’s and women’s
values differed more. Yet, Schwartz’s research and the World Values Survey show
unambiguously that these differences are not a function of MAS–FEM but of national
wealth, IDV–COLL and emancipation, which are unrelated to MAS–FEM (Minkov, 2018). In
wealthier, IDV countries, the self-reports of men and women produce larger differences in
values and personality traits between the genders, probably because the pressure for
conformism and cultural uniformity is lower than in COLL countries.
Minkov (2018) also provided primary data from a large cross-cultural study that analyzes
MAS and FEM self-construals, such as a desire to achieve fame and a desire to compete
(supposedly MAS traits as men are indeed more likely than women to describe themselves in
that way) vs a desire to help people (supposedly a FEM trait as women have a greater
tendency than men to report it). The analysis – with ipsatization or without it – did not
replicate MAS–FEM. Instead, these items were associated with IDV–COLL or MON–FLX, a
variant of Hofstede’s LTO dimension, described in the next section. However, Minkov’s (2018)
analysis simply examines associations between new measures of MAS–FEM items and the
old MAS–FEM index, without attempting to imitate Hofstede’s factor analysis, separating the
IDV–COLL variance from the MAS–FEM variance. The present study follows Hofstede’s
procedure more convincingly.

LTO theory and MON–FLX theory

These theories are relevant in our discussion of MAS–FEM because of the partial
conceptual overlap between them and the MAS–FEM dimension, which we targeted.
Inspired by Hofstede’s work, Canadian cross-cultural psychologist Michael Bond led
a team of researchers who used a list of Chinese values to compare national cultures
(Chinese Culture Connection, 1987). The study found that the East Asian countries
prioritized persistence (or perseverance), ordering relationships by status, thrift and a sense
of shame over personal steadiness and stability, protecting one’s face, respect for tradition
and reciprocation of greetings, favors and gifts. In Africa, Pakistan, the Philippines and the
English-speaking countries, it was the other way around: the second group of values was
prioritized over the first.
The relationships between these values seem enigmatic, yet Hofstede (2001) believed that
they depicted a contrast between future or long-term-oriented values (thrift and perseverance)
and past-and-present or short-term-oriented values (tradition and personal stability). In
Hofstede’s view, it is logical that countries that endorse the long-term values have
CCSM higher educational achievement in standardized international tests, as children in those
26,1 countries are socialized to prepare for the future, which is the main goal of modern education.
A focus on tradition and stability reveals a desire for preservation rather than positive
transformation, hence it accounts for lower educational achievement.
A study by Minkov and Hofstede (2012) replicated LTO with items from the World Values
Survey. However, it did not elucidate the complex associations between the dimension’s facets.
98 Minkov, Bond, Dutt, Schachner, Morales, Sanchez et al. (2018) proposed a reconceptualization of
this dimension and a new name (MON–FLX), whereas a study by Minkov, Dutt, Schachner,
Jandosova, Khassenbekov, Morales et al. (2018) shed more light on this dimension. From these
studies, we learn that MON societies (Latin America, Middle East and Africa) are characterized
by interdependence, involving an emphasis on mutual help and generosity. These acts, and the
recognition that people receive for them, seem to boost their self-regard and create a need to live
up to the positive image that one has earned. Therefore, it is important in those societies to
maintain one’s reputation by demonstrating personal stability: possession of strong values and
beliefs and always being the same admirable person who does not change much as a function of
shifting circumstances. Figuratively speaking, the human self should resemble a proud,
admirable and solid monument. FLX societies (East Asia, followed by Northern Europe)
emphasize self-reliance, hence mutual help and generosity are less important. Self-esteem is
lower than in MON societies and so is personal stability. It is important to be flexible and
adaptable rather than possess immutable values and beliefs.
MAS–FEM and MON–FLX overlap conceptually in two ways. First, people in MON societies
tend to be more willing to get involved in interpersonal competitions than those in FLX societies.
This is consistent with the MAS–FEM contrast that we know from Hofstede (2001): MON
societies appear to be MAS. Also, people in MON societies express a greater desire than those in
FLX societies to help other people and thus maintain good human relationships. This, however,
makes MON societies appear to be FEM. This discrepancy is puzzling, suggesting that either of
the two theories cannot be quite right. The present study addresses this vexing question.

Method
Questionnaire design
The questionnaire for this study was developed by the first author with the help of
consumer behavior and marketing research executives, such as the second author, and
consultants associated with MediaCom Ltd, a leading multinational media agency, which
also provided funding for this project. The questionnaire was tested and discussed with
consumer panels in the USA and Singapore. Preliminary versions were also fielded in
Kazakhstan and some Latin American countries. These consultations resulted in many
changes in the wordings of various original items and even item replacements.
As we were interested in the Hofstede model of national culture, our questionnaire design
started from a theoretical approach in an attempt to recreate Hofstede’s two original value
dimensions – IDV–COLL and MAS–FEM – which he obtained in the same factor analysis.
Following Hofstede’s (2001) theory, we assumed that people in so-called MAS societies
would value exhibition of personal wealth and competition for power. Therefore, they might
be likely to spend money on expensive products or lavish parties, as well as on seeking
political power. They might also be inclined to invest in business so as to earn even more
money. People in so-called FEM societies, which supposedly socialize for compassion and
concern for the environment, would be more likely to donate to hospitals and to the poor, as
well as for the preservation of nature. If people in MAS societies have an inclination to
donate at all, they might donate for the development of sports, since MAS involves approval
of competitions. Another cause for which they might donate is religion as Hofstede (2001)
believed that MAS cultures value religion more than FEM ones. However, as we formulated
these theoretical expectations, we were mindful of other perspectives as well. For instance, What people
Minkov et al. (2017) demonstrated statistically that religiousness is a pillar of COLL, which would do with
is unrelated to the MAS–FEM dimension, whereas Soyez (2012) found that an eco-friendly their money if
orientation is typical of IDV societies and is therefore unlikely to be associated with the
MAS–FEM dimension. It is also possible to theorize that donating for the development of they were rich
sports would be typical of COLL societies as sports boost group identities and collective
pride, or of MON because it is associated with a desire to compete. The analysis of our data 99
should resolve some of these controversies.
In addition to our theoretical approach, we also wanted our questionnaire to contain a
concise summary of all main types of spending choices that affluent people across the world
could make, as long as these choices could be expressed in a way that people in all cultures
find familiar and acceptable, so that cross-cultural comparisons can be made. These two
requirements for the questionnaire items – familiarity and acceptability – severely restricted
our options. For example, drawing up a list of expensive leisurely activities that most
ordinary people in all modern societies would be familiar with, and might wish to participate
in, proved challenging. Thus, practicing expensive sports was ruled out as we were advised
by some consultants that there are hardly large percentages of ordinary people in some
countries that have a good concept of them, let alone dream of becoming golfers or
yachtsmen. An item that asked the respondents if they would use much of their money to
attract lovers was dropped after some respondents in the pilot test deemed it offensive and
ethically unacceptable. Despite our belief that it is important to assess how much people
value education for themselves, there is no such item in the questionnaire as we realized that
while a good education costs a lot of money in some countries, it is quite inexpensive in other
societies due to state subsidies. Therefore, it would be misleading to ask people from diverse
countries if they would spend a lot of disposable income on obtaining a good education for
themselves and use their answers to compare values.
The final questionnaire asks the respondents whether, if they felt rich, they would save
large amounts of their wealth for bad times in the future, invest a lot so as to earn even
more, spend large amounts on self-boosting activities (buying expensive objects), on
hedonistic activities (travel), on activities that may be self-boosting and hedonistic at the
same time (lavish parties) and on power-seeking activities, as well as whether they would
donate large amounts for the preservation of nature, for science, education or art, for
religion, for hospitals and for the poor. The questionnaire is provided in Appendix 1 at the
end of this paper.
The questionnaire was translated into local languages except in a few countries where
English is an official language, although not necessarily a native one: India, Kenya, Nigeria,
Singapore and South Africa. In those countries, the English original was fielded, as high
percentages of respondents have university degrees and consequently a very good
command of English.

Samples
Data for this study were available from 51,529 respondents from 52 countries. Appendices 2
and 3 provide information about the composition of the samples. In 50 of the 52 countries,
the samples are consumer panels regularly used by the marketing research company
Lightspeed (lightspeedresearch.com). They are recruited probabilistically from the
population of each country, in an effort to obtain panels that reflect the national census
in terms of ethnic composition, geographic origin, age, gender, education and type of
employment. Our samples in the wealthy nations are close to this target. In the developing
nations, it proved difficult to reach respondents without higher education; therefore, most
respondents are university educated. In all other respects, the samples in most of the
developing nations are close to the national census. However, in a few multi-ethnic countries,
CCSM such as South Africa, the sample composition does not correspond closely to that of the
26,1 census in terms of ethnicity. This underrepresentation of some ethnic groups may not
matter much in countries such as Turkey, where the main ethnicities (Turks and Kurds)
have co-existed for thousands of years and may not have dramatically different values
today. Yet, we found that the responses of black and white South Africans differ
substantially. Therefore, we decided to use only the black sample as blacks account for more
100 than 80 percent of South Africa’s population.
In two countries, the data were not collected by Lightspeed because that company does
not have panels there. In Kazakhstan, the data were collected in face-to-face interviews by
Sange Inc., a sociological research agency. The sample provides complete and accurate
national representation for the purpose of a project for the Government of Kazakhstan. In
Peru, the data were provided by a local research agency. The sample consists mainly of
educated large city dwellers from Lima, Arequipa and Trujillo.
In all countries, except Kazakhstan, the data were collected online.
After analyzing our data in different ways, across subsamples with and without higher
education, we found that education differences do not affect factor structures and country
rankings. Therefore, we report data from analyses of entire samples.
As a first step, we calculated the percentages of respondents in each country that
selected “yes [I would use my money in that way]” and the percentages of respondents who
selected “maybe.” Assuming a 50-percent chance that respondents who select “maybe” will
use their money in the indicated way and a 50 percent chance that they will not, we divided
the percentage of each nation’s “maybe” responses by two and added the result to the
percentage of “yes” responses. Thus, the formula that we used to calculate each nation’s
score on each item was: percentage “yes” + 0.5 × percentage “maybe.” As a result, each
nation’s score is an estimate of the percentage of respondents who are likely to use their
money in the indicated way.
Our next concern was response bias as it was clear that respondents in some nations,
particularly those of Africa and Latin America, had a pronounced tendency to choose “yes”
responses, a phenomenon known as yea-saying or acquiescence, discussed in multiple
studies (Harzing, 2006; He et al., 2014; Johnson et al., 2005; Smith, 2004). Japanese
respondents exhibited the opposite tendency: to choose “no” regardless of the item content.
Taken at face value, both types of responses suggest unrealistic financial strategies. One is
unlikely at the same time to save, invest, spend and donate large amounts of one’s money for
very diverse causes. Also, it is impossible neither to save, nor to invest or spend. To obtain a
more realistic image of what a particular culture teaches people to do with their money, it is
necessary to assess how people collectively prioritize their financial strategies. In other
words, we need to understand the budget structure that a society as a whole would adopt.
This can be done by first adding up each nation’s scores on all items. That would
figuratively correspond to a nation’s total budget. Then, each raw item score is transformed
into a percentage of that budget. Statistically, when the raw scores are transformed in this
way, they are very similar to (in the sense of being strongly correlated with) ipsatized scores,
as used by Hofstede (2001) and Schwartz (1994, 2008a). Ipsatization, also known as
standardization within subject, is a calculation of the priority that a respondent or a whole
nation (depending on the level of analysis) ascribes to a particular value relative to other
values. Although its algorithm is different, the final result is very similar to a calculation of
percentages of a whole (the sum of all raw item scores) as performed in our study.

Results
A factor analysis across the 52 countries yielded four factors with eigenvalues over 1.00.
Yet, the eigenvalues of the third and fourth factors barely exceeded 1.00, and they explained
just a little more than 10 percent of variance each. Further, the fourth factor was defined
only by the item about donating to the poor. The third factor was defined by an What people
uninterpretable combination of items: donating for science and donating for sports vs would do with
saving. After dropping the item about donating to the poor and the one about donating for their money if
science, we obtained a fully interpretable two-factor structure with sufficiently high
loadings for all items. The two factors explained 70 percent of the variance. they were rich
We tried oblimin and promax rotations, with different κ and δ values, in order to obtain
moderately correlated factors as in Minkov (2018), who reported two main Hofstede 101
dimensions that are moderately correlated. However, all rotations yielded weakly correlated
factors. The highest correlation that we obtained was 0.34 after an oblimin rotation with the
δ value set at the highest possible level (0.8). We adopted that solution. Still, we note that all
other rotations yielded virtually identical factor structures and country positions. Therefore,
our findings would not have been substantially different if we had adopted a different type
of factor analysis. The adopted factor structure is provided in Table I. The two factors are
labeled altruism vs ego-boosting and hedonism vs conservatism. The term “egoism” for the
first factor was deliberately avoided as it tends to be equated with selfishness. However, a
willingness to give lavish parties does not suggest selfishness but an ambition to boost one’s
public image. The positive pole (corresponding to high scores) of the first factor could also
be called “ostentatiousness” as it reflects a desire to flaunt one’s wealth or an aspiration for
power, which also puts one in the public spotlight.
We provide country scores for the two dimensions in Tables II and III.
Figure 1 is a plot of the 52 countries on the two factors. Factor 2 scores were reversed by
multiplying them by −1, so as to align the corresponding axis with the North–South axis of
the Earth and with the IDV–COLL dimension whose scores descend from Northern Europe to
Africa and South Asia (Minkov et al., 2017). Figure 2 features the same plot but outlines eight
major cultural clusters. Essentially, Figures 1 and 2 are visualizations of Tables II and III. The
country plots are reminiscent of the geographic map of the world, as traditionally drawn in
European cartography, with the Americas to the West and East Asia to the East. The world
oceans are missing, which reflects the cultural proximity of parts of Southeast Asia, Africa
and Latin America, also known from other studies (Minkov et al., 2017; Minkov, Bond, Dutt,
Schachner, Morales, Sanchez et al., 2018, etc.). The resemblance between the real geographic
map of the world and the plots in Figures 1 and 2 demonstrates that our study has captured
some very real differences; otherwise it would be impossible to explain the geographic logic of
Figures 1 and 2. This validity criterion was borrowed from the work of Terracciano et al.
(2005) who argued that aggregate national measures of personality traits that do not yield
recognizable geographic patterns are illogical and must be flawed. Subsequently, Dobson and
Gelade (2012) showed that dimensions of national culture also have a tendency to yield

Loading on Factor 1: Loading on Factor 2:

ego-boosting (positive pole) vs conservatism (positive pole) vs
Item altruism (negative pole) hedonism (negative pole)

Would spend on expensive things 0.90 −0.07

Would spend on expensive parties 0.84 −0.22
Would donate to hospitals −0.83 0.20
Would use money to gain political power 0.79 0.01
Would invest in business −0.78 0.31
Would donate for religion −0.18 0.86
Would spend on travel −0.03 −0.85
Would donate for sports −0.45 0.73 Table I.
Would donate to preserve nature −0.11 −0.60 Factor structure
Would save 0.61 −0.69 across 52 nations
CCSM Japan 310
26,1 South Korea 229
China 174
Hong Kong 166
Germany 97
The Netherlands 87
France 87
102 Canada 83
Singapore 82
UK 79
Taiwan 79
Kazakhstan 79
Vietnam 55
India 55
Belgium 55
Australia 53
Thailand 51
Finland 48
Spain 35
Denmark 34
Sweden 33
USA 28
Austria 17
Switzerland 13
Russia 9
Italy 2
Ireland 1
Turkey −2
New Zealand −11
Poland −12
Malaysia −13
Indonesia −23
Norway −32
Greece −51
Czech Republic −55
Hungary −62
Brazil −65
Mexico −73
Serbia −81
Portugal −82
South Africa −85
Ukraine −87
Israel −89
The Philippines −109
Argentina −114
Table II. Peru −117
National scores for Romania −120
52 countries on Colombia −132
Factor 1: ego-boosting Chile −135
(high scores) vs Nigeria −154
altruism (low scores) Venezuela −164
( factor scores ×100) Kenya −171

geographic patterns. This has been used as a criterion to decide whether a particular measure
reflects a valid dimension of national culture or not (Minkov et al., 2013).
The cultural clusters in Figure 2 were drawn following the tradition of Inglehart and Baker
(2000): they are not an outcome of a clustering analysis but an attempt to outline continents
Nigeria 250
What people
Kenya 190 would do with
South Africa 177 their money if
Indonesia 148
Thailand 146
they were rich
The Philippines 127
Kazakhstan 127
Venezuela 118 103
Vietnam 114
Malaysia 104
Turkey 104
India 98
China 84
Hong Kong 55
Peru 53
Singapore 50
Colombia 48
Chile 44
Mexico 27
Taiwan 24
Brazil 16
Serbia 7
Ukraine 6
USA 2
Argentina 2
Romania 1
Poland −3
Israel −8
Italy −19
Ireland −36
Russia −37
Hungary −40
Canada −51
New Zealand −57
Australia −62
South Korea −63
Germany −68
Japan −77
Portugal −77
UK −80
Austria −88
Greece −91
Norway −104
Czech Republic −105
Spain −106
The Netherlands −106 Table III.
Belgium −111 National scores for
Switzerland −120 52 countries on
France −128 Factor 2: conservatism
Finland −129 (high scores) vs
Sweden −170 hedonism (low scores)
Denmark −188 ( factor scores ×100)

and civilizational conglomerates, such as Latin America, Anglo countries and Chinese-
speaking countries. Inglehart borrowed this approach from the work of Samuel Huntington.
One could prefer an empirical analysis, yet clustering tools can yield fairly different results
with respect to some cases, depending on the methodology that one decides to adopt.
CCSM 200.00
Denmark

26,1
Sweden

Finland
Hedonism

Switzerland Belgium
France
Czech Republic Norway
100.00 Spain The Netherlands
Austria
Greece
104 Portugal
New Zealand
Australia
UK
Germany
Korea
Japan

Hungary Canada
Ireland
Russia
Israel Italy
Argentina
USA
0.00 Serbia
Romania
Poland
Brazil
Ukraine
Colombia Mexico Taiwan

Chile
Peru Singapore Hong Kong

India
Turkey
China
–100.00
Malaysia Vietnam

Venezuela
The Philippines Kazakhstan

Indonesia Thailand
Conservatism

South Africa
–200.00 Kenya

Figure 1. Nigeria
A plot of 52 countries
on the two factors –300.00
of national culture
–200.00 –100.00 0.00 100.00 200.00 300.00 400.00
in this study
Altruism Ego-Boosting

Table IV provides validation for the Factors 1 and 2 scores through associations with
previously reported dimensions of national culture.
Table V provides evidence for some of the implications of the two factors. It shows how
they are associated with the results of a nationally representative study (International Social
Survey Program Research Group, 2009) of how often people attend sports events.

Discussion
Just like Hofstede’s analysis of 14 work goals, our 52-country study of 12 ways in which
people would use their disposable income yields two main dimensions of national culture,
underpinned by values. Conceptually, the first dimension resembles Hofstede’s MAS–FEM
as it contrast values that he considered MAS (status-seeking or ego-boosting) and values
that he considered FEM (compassion and maintenance of good human relationships).
Remarkably, we obtained this cultural contrast in a manner similar to Hofstede’s
statistical method. Like him, we did not analyze raw item scores but prioritizations: the
weights that a society ascribes to different types of values and the value ranking that these
weights produce.
However, the country index of this dimension is unrelated to Hofstede’s MAS–FEM. The
correlations that this dimension yields with previously reported dimensions in the literature
suggest unambiguously that it is yet another variant of MON–FLX (Minkov, Bond, Dutt,
Schachner, Morales, Sanchez et al., 2018), which is a reconceptualization of Confucian work
dynamism (Chinese Culture Connection, 1987) and LTO (Hofstede, 2001; Minkov and Hofstede,
2012). In other words, some of the values that Hofstede considered MAS, such as prioritization
of status-seeking or ego-boosting, are actually priorities in the FLX societies of East Asia, and,
 200.00 What people
would do with
their money if
Hedonism

they were rich

100.00 West Europe

Anglo Far East 105

South East Europe Russia

0.00

Latin America
Chinese

–100.00
South Asia
Conservatism

–200.00 Africa

Figure 2.
A plot of 52 countries
on the two factors of
national culture in this
–300.00
study, delineating
–200.00 –100.00 0.00 100.00 200.00 300.00 400.00
major cultural clusters
Altruism Ego-Boosting

Factor 1 Factor 2
Dimensions of national culture (ego-boosting) (conservatism) n (countries)

Flexibility–monumentalism (Minkov, Bond, Dutt, 0.88** 0.40** 52

Schachner, Morales and Sanchez, 2018)
Individualism–collectivism (Minkov et al., 2017) 0.42** −0.93** 52
Flexibility–monumentalism (Minkov, Dutt, Schachner, 0.81** 0.42** 52
Jandosova, Khassenbekov, Morales and Sanchez, 2018)
Individualism–collectivism (Minkov, Dutt, Schachner, 0.22 −0.90** 52
Jandosova, Khassenbekov, Morales and Sanchez, 2018)
Long-term orientation (Minkov and Hofstede, 2012) 0.85** 0.20 20
Embeddedness (Schwartz, 2008b) −0.32* 0.87** 49
Intellectual autonomy (Schwartz, 2008b) 0.25 −0.82** 49
Affective autonomy (Schwartz, 2008b) 0.35* −0.76** 49
Table IV.
Individualism–collectivism (Hofstede, 2001) 0.35* −0.67** 42
Correlations between
Power distance (Hofstede, 2001) 0.23 0.66** 42 the two factors in
Uncertainty avoidance (Hofstede, 2001) −0.17 0.09 42 this study and
Masculinity–femininity (Hofstede, 2001) 0.12 0.18 42 other dimensions of
Notes: *,**Correlation significant at 0.05 and 0.01, respectively national culture

to a lower extent, of some North European societies. Some of the values that Hofstede called
FEM, such as prioritizing compassion or altruism (being willing to donate for healthcare), are
actually priorities in MON societies. This is consistent with the findings of Minkov, Bond, Dutt,
Schachner, Morales, Sanchez et al. (2018): helping is a key MON value.
CCSM The conceptual association between our first dimension, ego-boosting vs altruism and
26,1 FLX–MON or LTO may not be glaringly evident due to a lack of a full overlap of the facets of
these dimensions. Yet, some facets do converge. We find that thrift has a low priority in
altruistic societies (Africa and Latin America), apparently because these are also MON
societies, socializing for helping and sharing. Thrift has a high priority in self-boosting
societies (East Asia) because these socialize for financial independence and self-sufficiency
106 (Minkov, Bond, Dutt, Schachner, Morales, Sanchez et al., 2018; Minkov, Dutt, Schachner,
Jandosova, Khassenbekov, Morales et al., 2018). This finding is also consistent with Confucian
work dynamism or LTO theory, which describes thrift as a key prioritization in East Asia but
not in Africa and Latin America (Chinese Culture Connection, 1987; Minkov and Hofstede,
2012). Also, from the perspective of MON–FLX theory, thrift is a dialectical ability to suppress
one’s natural desires and adopt a behavior that does not conform to them.
The finding in our study that the FLX societies of East Asia prioritize ostentatious ego-
boosting spending did not astonish the marketing and consumer behavior experts that helped
us with our work. Yet, the geographic contrasts that emerge from our study are somewhat
surprising. According to Schwartz (2008a, b), power, wealth and success, which he called
hierarchy values, are prioritized in East Asia, indeed, but also in Africa. We find that the three
African nations in our study – Nigeria, Kenya and South Africa – attach a lower importance to
ostentatious ego-boosting, relative to other values. There can be different explanations for this
discrepancy. Obviously, how people prioritize values depends on what comparisons they are
asked to make. Schwartz’s study finds that Africans attach a higher importance to hierarchy
values relative to egalitarian values. We do not have egalitarian values in our questionnaire as
these cannot be easily expressed as consumer choices. Instead, we see that in Africa some
types of compassion, without a close equivalent in Schwartz’s instrument, are deemed more
important than ego-boosting. Alternatively, it is possible that a desire to achieve success and
wealth, which is what Schwartz measured, and a desire to flaunt one’s success and wealth,
which is what we measured, are two different phenomena.
Why is ostentatious ego-boosting prioritized in FLX societies? We cannot be sure as this
is a novel finding. We can only surmise that it may be a negative reaction to the existing
cultural tendency to enforce a low self-esteem. Vice versa, as people in MON societies tend to
have high self-esteem, they may not feel a need to prioritize ostentatious ego-boosting.
The finding that a prioritization of ego-boosting, corresponding to the FLX pole of the
MON–FLX dimension, is negatively associated with interest in sports, as evident from the
factor structure and from the correlations in Table V, is logical as sports have a strong
competitive element that appeals to MON cultures, whereas people in FLX cultures are more
likely to state that they do not like to compete. It is harder to explain why prioritization of
ego-boosting (and FLX) is associated with a low prioritization of investment in business.

Factor 1 Factor 2 n
Variables (ego-boosting) (conservatism) (countries)

Percentage of respondents who attend sports events daily

(International Social Survey Program Research Group, 2009) −0.38 0.53* 26
Percentage of respondents who attend sports events several times a
week (International Social Survey Program Research Group, 2009) −0.50* 0.53* 26
Table V. Percentage of respondents who attend sports events several times a
Associations between
month (International Social Survey Program Research Group, 2009) −0.58* 0.32 26
the two factors in
this study and Percentage of respondents who attend sports events several
self-reported times a year or less (International Social Survey Program
frequencies of Research Group, 2009) 0.28 −0.55* 26
sports attendance Note: *Correlation significant at 0.01
This seems to reveal risk aversion in FLX societies, yet more research is needed to What people
understand this phenomenon properly. would do with
Our study confirms some well-known aspects of COLL and IDV societies and reveals new their money if
ones. It is not surprising that people in conservative-COLL societies have faith in religious
institutions and are willing to support them financially, or that IDV is associated with they were rich
hedonism, including travel for pleasure. It is also well-known that environmental concerns are
greater in IDV societies (Soyez, 2012). However, ours is the first large cross-cultural study 107
revealing that conservative-COLL societies prioritize sports, apparently because they boost
the collective identities of the fans, and collective identities are more important in COLL
societies than in IDV ones. The positive association between prioritization of saving and
hedonism-IDV is also a novel finding, awaiting a theoretical explanation.
In sum, our study confirms the need for a radical revision of Hofstede’s classic model of
national culture, reducing it to a two-dimensional model consisting of IDV–COLL and
MON–FLX. This model has emerged from studies of self-construals (Minkov, 2018),
parental advice to children (Minkov, Dutt, Schachner, Jandosova, Khassenbekov, Morales
et al., 2018) and now from value-related inclinations in consumer behavior. This does not
mean that one cannot or should not construct other dimensions of national culture and other
models. It simply means that, upon close scrutiny, these are the only two dimensions
in Hofstede’s model that are easily replicated, have sufficient internal consistency and
strong predictive properties with respect to external variables that withstand the effect of
plausible controls.
Finally, we must once again emphasize the fact that studying prioritizations, as in this
study, in Schwartz’s work, in Hofstede’s analysis of IDV–COLL and MAS–FEM work goals,
and in other studies, is not the same as studying raw scores: statements that have not been
ipsatized. For instance, East Asia scores high on thrift only after ipsatization, as in our study,
and the Chinese Culture Connection (1987). The World Values Survey section containing
desirable traits and values for children also suggests that thrift is highly valued in East Asia
but the respondents are asked to prioritize by selecting no more than five items out of ten.
This amounts to ipsatization. An analysis of our raw data indicates that importance of thrift is
associated with low national wealth: it is the African countries that have the highest scores.
However, we cannot know with certainty if this reflects an objective reality or is due to the fact
that the cultures of the developing nations socialize for acquiescence.

References
Chinese Culture Connection (1987), “Chinese values and the search for culture-free dimensions of
culture”, Journal of Cross-Cultural Psychology, Vol. 18 No. 2, pp. 143-164.
Crotts, J.C. and Erdman, R. (2000), “Does national culture influence consumers’ evaluation of travel
services? A test of Hofstede’s model of cross‐cultural differences”, Managing Service Quality,
Vol. 10 No. 6, pp. 410-419.
De Mooij, M. and Hofstede, G. (2002), “Convergence and divergence in consumer behavior: implications
for international retailing”, Journal of Retailing, Vol. 78 No. 1, pp. 61-69.
Dobson, P. and Gelade, G.A. (2012), “Exploring the roots of culture using spatial autocorrelation”,
Cross-Cultural Research, Vol. 46 No. 2, pp. 160-187.
Dwyer, S., Mesak, H. and Hsu, M. (2005), “An exploratory examination of the influence of national
culture on cross-national product diffusion”, Journal of International Marketing, Vol. 13 No. 2,
pp. 1-27.
Gong, W. (2009), “National culture and global diffusion of business-to-consumer e-commerce”, Cross
Cultural Management, Vol. 16 No. 1, pp. 83-101.
Harzing, A.W. (2006), “Response styles in cross-national survey research: a 26-country study”,
International Journal of Cross-Cultural Management, Vol. 6 No. 2, pp. 243-266.
CCSM He, J., Bartram, D., Inceoglu, I. and van de Vijver, F.J.R. (2014), “Response styles and personality traits:
26,1 a multilevel analysis”, Journal of Cross-Cultural Psychology, Vol. 45 No. 7, pp. 1028-1045.
Henry, W.A. (1976), “Cultural values do correlate with consumer behavior”, Journal of Marketing
Research, Vol. 13 No. 2, pp. 121-127.
Hofstede, G. (2001), Culture’s Consequences: Comparing Values, Behaviors, Institutions, and
Organizations Across Nations, 2nd ed., Sage, Thousand Oaks, CA.
108 House, R.J., Hanges, P.J., Javidan, M., Dorfman, P.W. and Gupta, V. (Eds) (2004), Culture, Leadership,
and Organizations: The GLOBE Study of 62 Societies, Sage, Thousand Oaks, CA.
Inglehart, R. and Baker, W.E. (2000), “Modernization, cultural change, and the persistence of traditional
values”, American Sociological Review, Vol. 65 No. 1, pp. 19-51.
International Social Survey Program Research Group (2009), “International Social Survey Program:
Leisure Time and Sports – ISSP 2007”, ZA4850 data file version 2.0, GESIS Data Archive,
Cologne, doi: 10.4232/1.10079.
Johnson, T., Kulesa, P., Cho, Y.I. and Shavitt, S. (2005), “The relation between culture and response styles:
evidence from 19 countries”, Journal of Cross-Cultural Psychology, Vol. 36 No. 2, pp. 264-277.
Kacen, J.J. and Lee, J.A. (2002), “The influence of culture on consumer impulsive buying behavior”,
Journal of Consumer Psychology, Vol. 12 No. 2, pp. 163-176.
Kluckhohn, C. (1967), “Values and value-orientations in the theory of action: an exploration in definition
and classification”, in Parsons, T. and Shils, E.A. (Eds), Toward a General Theory of Action,
Harvard University Press, Cambridge, MA, pp. 388-433.
Liu, R.R. and McClure, P. (2001), “Recognizing cross‐cultural differences in consumer complaint
behavior and intentions: an empirical examination”, Journal of Consumer Marketing, Vol. 18
No. 1, pp. 54-75.
Minkov, M. (2018), “A revision of Hofstede’s model of national culture: old evidence and new data from
56 countries”, Cross-Cultural and Strategic Management, Vol. 25 No. 2, pp. 231-256.
Minkov, M. and Hofstede, G. (2012), “Hofstede’s fifth dimension; new evidence from the world values
survey”, Journal of Cross-Cultural Psychology, Vol. 43 No. 1, pp. 3-14.
Minkov, M., Blagoev, V. and Hofstede, G. (2013), “The boundaries of culture; do questions about
societal norms reveal cultural differences?”, Journal of Cross-Cultural Psychology, Vol. 44 No. 7,
pp. 1094-1106.
Minkov, M., Bond, M.H., Dutt, P., Schachner, M., Morales, O., Sanchez, C.J., Jandosova, J.,
Khassenbekov, Y. and Mudd, B. (2018), “A reconsideration of Hofstede’s fifth dimension: new
flexibility versus monumentalism data from 54 countries”, Cross-Cultural Research, Vol. 52
No. 3, pp. 309-333.
Minkov, M., Dutt, P., Schachner, M., Jandosova, J., Khassenbekov, Y., Morales, O. and Sanchez, C.J.
(2018), “What values and traits do parents teach their children? New data from 54 countries”,
Comparative Sociology, Vol. 17 No. 2, pp. 221-252.
Minkov, M., Dutt, P., Varma, T., Schachner, M., Morales, O., Sanchez, C.J. et al. (2017), “A revision of
Hofstede’s individualism-collectivism dimension: a new national index from a 56-country study”,
Cross-Cultural and Strategic Management, Vol. 24 No. 3, pp. 386-404.
Moon, J., Chadee, D. and Tikoo, S. (2008), “Culture, product type, and price influences on consumer
purchase intention to buy personalized products online”, Journal of Business Research, Vol. 61
No. 1, pp. 31-39.
Pavlou, P.A. and Chai, L. (2002), “What drives electronic commerce across cultures? A cross-cultural
empirical investigation of the theory of planned behavior”, Journal of Electronic Commerce
Research, Vol. 3 No. 4, pp. 240-253.
Radojevic, T., Stanisic, N. and Stanic, N. (2017), “Inside the rating scores: a multilevel analysis of the
factors influencing customer satisfaction in the hotel industry”, Cornell Hospitality Quarterly,
Vol. 58 No. 2, pp. 134-162.
Schwartz, S.H. (1994), “Beyond individualism/collectivism: new cultural dimensions of values”, in Kim, U., What people
Kagitcibasi, C., Triandis, H.C., Choi, S.C. and Yoon, G. (Eds), Individualism and Collectivism: would do with
Theory, Method, and Application, Sage, Thousand Oaks, CA, pp. 85-119.
their money if
Schwartz, S.H. (2008a), Cultural Value Orientations: Nature and Implications of National Differences,
State University – Higher School of Economics Press, Moscow. they were rich
Schwartz, S.H. (2008b), “The seven Schwartz cultural value orientations for 80 countries”,
ResearchGate, available at: www.researchgate.net/publication/304715744_The_7_Schwartz_ 109
cultural_value_orientation_scores_for_80_countries (accessed September 2018).
Schwartz, S.H. and Bardi, A. (2001), “Value hierarchies across cultures: taking a similarities
perspective”, Journal of Cross-Cultural Psychology, Vol. 32 No. 3, pp. 268-290.
Smith, P.B. (2004), “Acquiescence response bias as an aspect of cultural communication style”, Journal
of Cross-Cultural Psychology, Vol. 35 No. 1, pp. 50-61.
Soyez, K. (2012), “How national cultural values affect pro-environmental consumer behavior”,
International Marketing Review, Vol. 29 No. 6, pp. 623-646.
Swidler, A. (1986), “Culture in action: symbols and strategies”, American Sociological Review, Vol. 51
No. 2, pp. 273-286.
Terracciano, A., Abdel-Khalek, A.M., Adam, N., Adamovova, L., Ahn, C.K., Ahn, H.N. et al. (2005),
“National character does not reflect mean personality trait levels in 49 cultures”, Science, Vol. 310
No. 5745, pp. 96-100.
Tung, R.L. and Stahl, G. (2018), “The tortuous evolution of the role of culture in IB research: What we
know, what we don’t know, and where we are headed”, Journal of International Business Studies,
Vol. 49 No. 9, pp. 1167-1189.
Yeniyurt, S. and Townsend, J. (2003), “Does culture explain acceptance of new products in a country?
An empirical investigation”, International Marketing Review, Vol. 20 No. 4, pp. 377-396.

Appendix 1. Questionnaire
If I were rich:
• VR1. I would spend much of my money on expensive things to show how successful I am.
• VR2. I would save much of my money for bad times in the future.
• VR3. I would spend much of my money on large and expensive parties and celebrations.
• VR4. I would donate much of my money to poor people.
• VR5. I would invest much of my money in business that will help me earn even more.
• VR6. I would donate much of my money for the development of science, education, or art.
• VR7. I would use much of my money to gain political power.
• VR8. I would use much of my money to travel and see the cultures, art, or nature of countries
on all continents.
• VR9. I would donate much of my money to preserve nature and wild animals.
• VR10. I would donate much of my money to promote some sports.
• VR11. I would donate much of my money to religious organizations.
• VR12. I would donate much of my money to hospitals or other health organizations.
The response options for all items are:
(1) Yes
(2) Maybe
(3) No
CCSM Appendix 2
26,1

Language of the n respondents and gender Education: % without higher

Country questionnaire (n men−n women) education−% with higher education

110 Argentina Spanish 505 (245–260) 40–60

Australia English 690 (348–342) 53–47
Austria German 251 (118–133) 77–33
Belgium Dutch, French 1,021 (508–513) 53–47
Brazil Brazilian Portuguese 1,524 (726–798) 33–67
Canada English, French 517 (257–260) 40–60
Chile Spanish 522 (190–332) 49–51
China Chinese 2,043 (1,001–1,042) 8–92
Colombia Spanish 509 (260–249) 24–76
Czech Republic Czech 564 (298–266) 78–22
Denmark Danish 256 (118–138) 48–52
Egypt Arabic 540 (314–226) 1–99
Finland Finnish 252 (124–128) 59–41
France French 1,037 (542–496) 59–41
Germany German 1,037 (471–566) 79–21
Greece Greek 647 (249–398) 22–78
Hong Kong Chinese 256 (116–140) 41–59
Hungary Hungarian 507 (246–261) 53–47
India English 2,028 (1,063–965) 10–90
Indonesia Indonesian 2,042 (1,095–947) 21–79
Ireland English 265 (128–137) 56–44
Israel Hebrew 365 (196–169) 44–56
Italy Italian 1,075 (568–507) 63–27
Japan Japanese 1,523 (1,004–519) 40–60
Kazakhstan Russian and Kazakh 8,478 (3,981–4,497) 91–8
Kenya English 518 (266–252) 30–70
Malaysia Malaysian and English 769 (387–382) 37–63
Mexico Spanish 1,521 (782–739) 25–75
The Netherlands Dutch 567 (249–318) 50–50
New Zealand English 321 (151–170) 50–50
Nigeria English 1,243 (718–525) 20–80
Norway Norwegian 258 (126–132) 55–45
Peru Spanish 537 (219–318) 30–70
The Philippines Filipino/Tagalog 1,521 (594–927) 10–90
Poland Polish 506 (199–307) 40–60
Portugal Portuguese 504 (269–235) 47–53
Romania Romanian 534 (286–248) 34–66
Russia Russian 1,597 (736–861) 21–79
Serbia Serbian 248 (116–132) 54–46
Singapore English 403 (209–194) 47–53
South Africa English 1,013 (504–509) 60–40
South Korea Korean 1,033 (510–523) 22–78
Spain Spanish 505 (251–254) 39–61
Sweden Swedish 687 (294–393) 65–35
Table AI. Switzerland German, French 252 (118–134) 72–28
Sample characteristics: Taiwan Chinese 553 (292–261) 20–80
language of the Thailand Thai 1,032 (504–528) 13–87
questionnaire, number Turkey Turkish 1,095 (546–549) 26–74
of respondents, gender UK English 1,522 (750–772) 57–43
balance self-reported Ukraine Ukrainian 510 (258–252) 30–70
race or ethnicity and USA English 2,467 (1,188–1,279) 44–56
geographic origin of Venezuela Spanish 509 (254–255) 22–78
the respondents Vietnam Vietnamese 1,028 (560–469) 17–83
Appendix 3 What people
would do with
their money if
Geographic origin they were rich
A: self-reported towns or regions where the
Self-reported race or ethnicity ( for respondents grew up until age 18 (%)
Canada, Belgium, India, Israel, the or 111
Philippines, Spain, Switzerland: language B: research agency data: towns or regions where
Country spoken at home as a child) (%) the respondents were recruited (%)

Australia Unspecified B:
New South Wales ¼ 30%
Victoria ¼ 26%
Queensland ¼ 22%
Western Australia ¼ 9%
South Australia ¼ 8%
Capital Territory ¼ 2%
Tasmania ¼ 2%
Northern Territory ¼ 1%
Belgium Dutch ¼ 57% B:
French ¼ 38% Antwerp ¼ 15%
Arabic ¼ 1% East Flanders ¼ 14%
Turkish ¼ 1% Hainaut ¼ 14%
Other ¼ 3% Liege ¼ 12%
West Flanders ¼ 10%
Flemish Brabant ¼ 9%
Brussels ¼ 8%
Limburg ¼ 7%
Walloon Brabant ¼ 4%
Namur ¼ 4%
Luxemburg ¼ 3%
Brazil White ¼ 76% A:
Black ¼ 12% Sao Paolo ¼ 36%
Amerindian ¼ 1% Rio de Janeiro ¼ 19%
Mixed/Other ¼ 11% Rio Grande Sul ¼ 9%
Minas Gerais ¼ 8%
Parana ¼ 7%
Bahia ¼ 5%
Pernambuco ¼ 3%
Goias ¼ 3%
Other ¼ 10%
Canada Language: B:
English ¼ 72% Ontario ¼ 40%
French ¼ 28% Quebec ¼ 23%
Race/Ethnicity: Prairies ¼ 17%
White ¼ 79% British Columbia ¼ 12%
East Asian ¼ 9% Atlantic ¼ 9%
South Asian ¼ 4%
Black ¼ 3%
Other ¼ 5% Table AII.
China Unspecified A: Sample characteristics
Shanghai ¼ 19% in large or multi-
Guangdong ¼ 14% ethnic countries: self-
Beijing ¼ 12% reported race,
Hubei ¼ 7% ethnicity, or native
language and
geographic origin of
(continued ) the respondents
CCSM
26,1

Geographic origin
A: self-reported towns or regions where the
Self-reported race or ethnicity ( for respondents grew up until age 18 (%)
112 Canada, Belgium, India, Israel, the or
Philippines, Spain, Switzerland: language B: research agency data: towns or regions where
Country spoken at home as a child) (%) the respondents were recruited (%)

Sichuan ¼ 5%
Zhejiang ¼ 5%
Shandong ¼ 5%
Liaoning ¼ 4%
Fujian ¼ 3%
Anhui ¼ 3%
Henan ¼ 3%
Tianjin ¼ 3%
Hunan ¼ 2%
Jiangxi ¼ 2%
Other ¼ 4%
Colombia Unspecified A:
Bogota, Centro ¼ 53%
Medellin-Antioquia ¼ 17%
Cali, Pacifica ¼ 11%
Barranquilla, Caribe ¼ 10%
Oriente ¼ 4%
Other ¼ 5%
Egypt Unspecified A:
Cairo region ¼ 58%
Alexandria/Matrouh ¼ 21%
Sinai ¼ 19%
Other ¼ 2%
France Unspecified A:
Paris ¼ 21%
Nord ¼ 10%
Est ¼ 14%
Ouest ¼ 16%
Sud-Ouest ¼ 12%
Sud-Est ¼ 12%
Mediterranee ¼ 8%
Other ¼ 7%
Germany German ¼ 94% A:
Russian ¼ 2% West Germany ¼ 72%
Turkish ¼ 1% East Germany ¼ 22%
Other ¼ 3% Other ¼ 6%
India Hindi ¼ 33% A:
English ¼ 13% Delhi ¼ 15%
Tamil ¼ 10% Mumbai ¼ 14%
Bengali ¼ 8% Kolkata ¼ 8%
Telugu ¼ 7% Chennai ¼ 7%
Gujarati ¼ 6% Tamil Nadu ¼ 6%
Marathi ¼ 6% Bangalore ¼ 5%
Malayalam ¼ 5% Hyderabad ¼ 5%
Kannada ¼ 4% Uttar Pradesh ¼ 4%

Table AII. (continued )

 What people
would do with
their money if
Geographic origin they were rich
A: self-reported towns or regions where the
Self-reported race or ethnicity ( for respondents grew up until age 18 (%)
Canada, Belgium, India, Israel, the or 113
Philippines, Spain, Switzerland: language B: research agency data: towns or regions where
Country spoken at home as a child) (%) the respondents were recruited (%)

Urdu ¼ 2% Andhra Pradesh ¼ 4%

Other ¼ 6% Ahmedabad ¼ 3%
Kerala ¼ 3%
Rajasthan ¼ 2%
Karnataka ¼ 2%
Gujarat ¼ 2%
Maharashtra-Pune ¼ 2%
Other ¼ 18%
Indonesia Javanese ¼ 52% Unspecified
Chinese ¼ 16%
Sundanese ¼ 11%
Batak ¼ 5%
Malay ¼ 4%
Betawi ¼ 3%
Minangkabau ¼ 3%
Buginese ¼ 2%
Madurese ¼ 1%
Other ¼ 3%
Israel Hebrew ¼ 81% A:
Russian ¼ 7% Grew up in Israel ¼ 83%
Arabic ¼ 2% Immigrant ¼ 17%
Other ¼ 10%
Italy Italian ¼ 98% A:
Other ¼ 2% Southern Italy ¼ 43%
Northern Italy ¼ 37%
Central Italy ¼ 19%
Other ¼ 1%
Japan Unspecified B:
The sample composition approximates that
of the census: 80 percent from Chubu, Kanto, and
Kinki and 20 percent from Tohoku, Chugoku,
Kyushu and Shikoku
Kazakhstan Kazakh ¼ 65% B:
Russian ¼ 21% Proportionately sampled from all 16
Uzbek ¼ 4% administrative regions of Kazakhstan, based on
Ukrainian ¼ 2% their population
Other ¼ 8%
Kenya Kikuyu ¼ 31% Unspecified
Luo ¼ 19%
Luhya ¼ 14%
Kamba ¼ 10%
Kalenjin ¼ 9%
Other ¼ 17%
Malaysia Malay ¼ 48% A:
Chinese ¼ 45% Selangor ¼ 21%

(continued ) Table AII.

CCSM
26,1

Geographic origin
A: self-reported towns or regions where the
Self-reported race or ethnicity ( for respondents grew up until age 18 (%)
114 Canada, Belgium, India, Israel, the or
Philippines, Spain, Switzerland: language B: research agency data: towns or regions where
Country spoken at home as a child) (%) the respondents were recruited (%)

Indian ¼ 5% Kuala Lumpur ¼ 15%

Other ¼ 2% Perak ¼ 12%
Johor ¼ 12%
Penang ¼ 10%
Sarawak ¼ 6%
Pahang ¼ 5%
Negeri Sembilan ¼ 3%
Melaka ¼ 3%
Sabah ¼ 3%
Kelantan ¼ 3%
Terengganu ¼ 2%
Other ¼ 5%
Mexico Unspecified B:
The sample composition approximates that of
the census: 20 percent from Mexico City and
Distrito Federal, 30 percent from Nuevo Leon,
Jalisco, Veracruz, Puebla, Baja California,
Guanajuato, Chihuahua and Tamaulipas, 50
percent from diverse unspecified regions
Nigeria Yoruba ¼ 41% A:
Igbo ¼ 29% Southern Nigeria ¼ 66%
Hausa ¼ 2% Northern Nigeria ¼ 15%
Other ¼ 18% Unspecified ¼ 19%
The Tagalog ¼ 53% A:
Philippines Bisaya ¼ 12% Manila region ¼ 44%
Cebuano ¼ 11% Balance Luzon ¼ 21%
English ¼ 6% Visyas ¼ 18%
Hiligaynon ¼ 5% Mindanao ¼ 12%
Ilocano ¼ 4% Other ¼ 5%
Bicol ¼ 4%
Chinese ¼ 1%
Pangasinan ¼ 1%
Other ¼ 3%
Russia A:
Moscow ¼ 20%
Volga region ¼ 20%
Siberia ¼ 11%
Ural ¼ 8%
Saint Petersburg ¼ 6%
Far East ¼ 4%
Other ¼ 31%
Singapore Chinese ¼ 90% Unspecified
Malay ¼ 5%
Indian ¼ 3%
Other ¼ 2%

Table AII. (continued )

 What people
would do with
their money if
Geographic origin they were rich
A: self-reported towns or regions where the
Self-reported race or ethnicity ( for respondents grew up until age 18 (%)
Canada, Belgium, India, Israel, the or 115
Philippines, Spain, Switzerland: language B: research agency data: towns or regions where
Country spoken at home as a child) (%) the respondents were recruited (%)

South Africa Self-reported: Unspecified

Black(Zulu) ¼ 10%
Black(Sotho) ¼ 5%
Black(Xhosa) ¼ 5%
Black(Tswana) ¼ 3%
Black(Venda) ¼ 2%
Black(Swazi) ¼ 1%
Black(Ndebele) ¼ 1%
White(English) ¼ 33%
White Afrikaner ¼ 25%
Indian ¼ 8%
Malay or Indonesian ¼ 1%
Mixed or other ¼ 7%
South Korea Unspecified According to Lightspeed GMI data, the sample
composition approximates that of the census: 60
percent from Greater Seoul, 40 percent from
Daegu and Kyungbuk, Daejeon and
Chungcheong, as well as Gwangju, Jeolla and
Kwangwon
Spain Castilian ¼ 83% Self-reported:
Catalan ¼ 9% Andalucia ¼ 19%
Galician ¼ 2% Catalunya ¼ 15%
Other ¼ 6% Castilla y Leon ¼ 8%
Galicia ¼ 7%
Euskadi (Pais Vasco) ¼ 7%
Islas ¼ 4%
Other ¼ 40%
Switzerland German ¼ 55% According to Lightspeed GMI:
French ¼ 32% Lake Leman/Geneva ¼ 33%
Italian ¼ 4% Central ¼ 26%
Other ¼ 9% Zurich ¼ 13%
Northwest ¼ 10%
East ¼ 10%
Ticino ¼ 2%
Thailand Unspecified B:
Approximately 50 percent from the region of
Bangkok, 50 percent from Pattaya-Chon Buri,
Chiang Mai, Hat-Yai Songkhla and Nakhon
Ratchasima
Turkey Turk ¼ 96% A:
Kurd ¼ 2% Marmara ¼ 34%
Other ¼ 2% Anatolia ¼ 16%
Aegean coast ¼ 13%
Mediterranean coast ¼ 10%
Ankara ¼ 9%

(continued ) Table AII.

CCSM Geographic origin
26,1 A: self-reported towns or regions where the
Self-reported race or ethnicity ( for respondents grew up until age 18 (%)
Canada, Belgium, India, Israel, the or
Philippines, Spain, Switzerland: language B: research agency data: towns or regions where
Country spoken at home as a child) (%) the respondents were recruited (%)

116 Black Sea coast ¼ 8%

Istanbul ¼ 8%
Izmir ¼ 2%
Other ¼ 1%
UK Unspecified A:
South England ¼ 44%
North England ¼ 33%
Scotland ¼ 8%
Wales ¼ 5%
Northern Ireland ¼ 1%
Other ¼ 9%
USA White ¼ 75% A:
Black ¼ 9% Midwest ¼ 27%
Hispanic ¼ 7% Northeast ¼ 26%
Asian ¼ 7% Southeast ¼ 19%
Other ¼ 2% West ¼ 18%
Southwest ¼ 8%
Other ¼ 2%
Note: In some columns, the sum of percentages may exceed 100 as some respondents may have provided
Table AII. more than one answer

Other sample characteristics

In each country, approximately 90 percent of all respondents are more or less equally distributed
across a specific age range. In countries with older populations, such as those of Europe, this age range
is between 18–20 and 70. In continents and countries with young populations, such as those of Africa
and India, the typical age range is between 18–20 and 40 or 18–20 and 50. There are some exceptions to
this rule, such as Italy (an aging population) where the age range is mostly between 20 and 55.
The respondents were asked to provide only a very broad indication of the type of job level that
they held at the moment of the survey: senior, mid-level or junior manager or expert, skilled manual
worker or semi-skilled manual worker. The other options were “student,” “unemployed” and “retired.”
Since the respondents in the developing countries have higher educational degrees, they are more
likely to hold senior or mid-level manager or expert positions: typically between 30 and 40 percent of
the sample. In developed countries, this percentage typically ranges between 15 and 35. Percentages of
junior managers or experts are between 15 and 30 in most countries, whereas those of workers are
20–25 percent in the developed countries but below 10 percent in developing countries, such as Nigeria
and India. Percentages of students range between 5 and 25 in most countries, whereas those of
unemployed people are roughly between 5 and 15 percent. The percentages of retirees are small,
typically below 10 percent in all countries. No questions were asked about the type of industry or
service in which the respondents are employed but, according to Lightspeed, these are quite diverse.

Corresponding author
Michael Minkov can be contacted at: [email protected]

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: [email protected]