A longitudinal study of cultural differences in global corporate ...

Journal of International Business and Cultural Studies
A longitudinal study of cultural differences in global
corporate web sites

Stephanie S. Robbins
UNC – Charlotte

Antonis C. Stylianou
UNC – Charlotte

Abstract

Ten years ago, during the early days of the Internet, research was undertaken that looked
for evidence of cultural convergence in the web sites of global corporations using Hofstede’s
cultural dimensions and country clusters. That study found little homogeneity. Since then, the
Internet has matured and corporations have become more sophisticated in designing their web
sites. Therefore, the main objective of the current project is to provide a longitudinal view of
how national culture has changed over a ten year period by replicating the 1998 study and
comparing the results. Findings suggest that there have been some changes towards
homogeneity.

Keywords: Cultural Convergence; Internet Homogenizing Effect; Standardization; Global
Corporations; Hofstede


A Longitudinal Study of Cultural, Page 1

---

Journal of International Business and Cultural Studies
INTRODUCTION

Over the last ten years (1998-2008), Internet usage has grown from 147 million
(3.6%) worldwide users to 1,319 million (20%) (Group, 2008). In looking over
worldwide usage of the Internet, 1999 seems to represent a turning point because US
Internet users no longer represented a majority of users (Shirky, 2008). This can be
attributed to a number of factors, including the declining cost of technology, which made
it more affordable to larger segments of humanity (both in developed as well as
developing countries) (Akmanligil, & Palvia, 2004), increases in network bandwidth, and
the advances in wireless networks, which have allowed corporations to extend their reach
beyond national borders. The resulting globalization has changed the way corporations
operate and conduct business. The use of technology has also allowed companies not
previously competitive to successfully operate in a multinational environment. But
globalization has also created challenges (Akmanligil, & Palvia, 2004) because it is more
difficult to balance the needs of local markets, which have specific language preferences
and cultural identities, with the need to create a cohesive global corporate image.
Companies moving into the global arena find that traditional methods of tailoring
advertising campaigns, product mixes, and brand image to reflect the tastes and cultural
patterns of specific countries or regions do not always work. A successful web site
designed for the global marketplace must be understandable by a diverse group of
individuals (Radzievshy, 1998) yet somehow have the flexibility to offer language
options and capture the cultural norms of each country (Horwitt, 1997; Horwitt, 1997).
Given the rapid acceptance of the Internet by people around the world, one must
question what kind of influence the Internet has had on national culture. Corporate
culture has become increasingly visible and can be seen through the way a company
designs its web site (Want, 2003). Thus, one can readily observe and compare
differences in global corporate culture. The main objective of this study is to update
research that was undertaken in the early days of the Internet (Robbins, & Stylianou,
2001-2002, Winter). This project takes a longitudinal view to determine if differences in
national culture found in 1998 (Robbins, & Stylianou, 2001-2002, Winter) are still
evident in the web sites of global corporations or if national cultures, as reflected through
these web sites, have converged.

LITERATURE REVIEW

Culture can be defined as a shared set of values that influence societal
perceptions, attitudes, preferences, and responses (Hofstede, 1980.) Today these
characteristics are easily visible in the marketplace (Want, 2003). One of the major
components of culture is language (Hofstede, 2001). English, although the mother tongue
to people in only six countries (dara.NET, No date), has been and still is the dominant
language of the Internet ( Group, 2008). However, today there is a marked increase in the
use of other languages, including Chinese, Spanish, and Japanese (Group, 2008), and
customer-focused online language translation services are now available. The increase in
the number of people preferring to use languages other than English online will have a
profound implication on how the Internet is used and how web sites are developed. As
the Internet becomes more of a global mass medium, the way commerce, news, research
A Longitudinal Study of Cultural, Page 1

---

Journal of International Business and Cultural Studies
and entertainment are presented has had to be rethought. For example, Japanese
companies understood the need for the availability of language options early and so now
routinely provide both Japanese and English versions of their sites that can be accessed
with a single click. In fact, several of the major search engines are offering their services
in multiple languages. Furthermore, Internet providers are recognizing that as their
audience broadens, so must their language support. It is an open debate whether the
English language will continue to be the standard language of cyberspace, or if it will
deepen its hold on the world as more people go online, or even if its usage will decline
and it will not even survive as the world’s lingua franca. Representing one side of the
debate, Pinker (2002, January 2) argues that “the Internet is [already] becoming polyglot,
and with improving translation engines, English will not be indispensable much longer on
the Web.” So the question becomes: How has the use of language on global web sites
changed over the last ten years?
Other cultural dimensions have been identified and over the years a number of
theoretical models have been presented in the literature that allow for the measurement
and comparison of different cultures based on a number of factors (for overviews see
(Adler, Koktor, & Redding, 1986; Ronen, & Shenkar, 1985). For example, Hall (1976)
presented a four factor model that distinguished cultures based on communication,
Hofstede developed a five factor model that will be discussed in depth in a later section
of this paper, and Trompenaars (1993) developed a seven-factor values model. Schwartz
(1984) also developed a seven-factor model along the lines of Hofstede and, finally,
Khaslavsky designed a nine-factor model based on the work of the previously mentioned
authors. Smith, Dunckley, French (2004) introduced the concept of cultural attractors
such as the use of color and color combination, trust signs, and the use of metaphor and
language cues. However, it is not clear how some of these cultural attractors were
operationalized. Badre (2002) approached the problem from a cultural marker
perspective that included such factors as color, spatial organization, fonts, icons, and
metaphors.
While there has been a lot of work done in the area of identifying cultural factors,
the model that is most often cited and accepted (Choe, 2004; Kondratove, Goldfarb,
Gervais, & Fournier, 2005; Marcus, & Gould, 2000; Robbins, & Stylianou, 2003; Singh,
& Baack, 2004; Singh, Kumar, & Baack, 2005; Sondergaard, 1994) is the typology of
culture developed by Hofstede (1980). His model is based on an extensive survey of
foreign subsidiaries of IBM and Hofstede concluded that national cultures differ along
five dimensions: power distance, uncertainty avoidance, individualism, masculinity, and
long-term orientation. These dimensions will be discussed in depth in the methodology
section of this paper. Hofstede also categorized countries based on commonalities and
similar histories. This resulted in the identification of six clusters: Anglo, Nordic,
German, Latin, Asian, and, Japan.
Even though Hofstede’s cultural dimensions model is widely used by researchers
today, his typology was based on research that was concluded thirty years ago.
Therefore, one must wonder if his hypotheses are still valid. Hofstede argued that
cultural values are deeply ingrained, enduring, and long-lasting. If that is true, then
Hofstede’s model retains validity. Also, there have been subsequent studies that provide
evidence that Hofstede’s cultural dimensions are still relevant (Adler, Koktor, &
Redding, 1986; Hoppe, 1992; Robbins, & Stylianou, 2001-2002, Winter; Shackleton, &
A Longitudinal Study of Cultural, Page 2

---

Journal of International Business and Cultural Studies
Ali, 1990; Singh, & Baack, 2004). However, there are a number of other studies that
have reached a different conclusion (Levitt, 1983; Nordstrom, 1991; O’Reilly, 1991;
Ohmae, 1985; Roberts, & Boyacigille, 1984; Triandis, 1982). This body of research
suggests that core cultural values have converged thus making Hofstede’s categorization
irrelevant.
It is thought that globalization supported by the Internet and other technologies
has had a homogenizing effect that has led to this cultural convergence (CPSR, 1997;
Deen, 1999; Lee, 1998; Rothkopf, 1997). And, it could be argued that the world
consumption of Coca-Cola and McDonald’s hamburgers are examples of such
convergence. If homogenization has taken place, then there should be evidence of this
effect in the web pages of global corporations. However, research undertaken in the early
days of the Internet did not provide support for that (Robbins, & Stylianou, 2001-2002,
Winter). Using Hofstede’s typology, that study examined the web sites of 90 global
corporations and found that the Internet had had some impact on culture but found no
evidence of cultural convergence at that time (Robbins, & Stylianou, 2001-2002, Winter).
That study was undertaken ten years ago and much has happened in the world since then,
including the maturing of the Internet and the growing sophistication of global
corporations in designing their web sites. Today, the question of cultural convergence is
still not clear. To understand any changes that might have happened over this time period,
it would be necessary to undertake a longitudinal study to see if shifts in cultural patterns
are evident. To date, no research has been undertaken that looks at web cultural
dimensions over time. This study has been designed to address this void in the literature.

METHODOLOGY

This project was developed to reflect Hofstede’s (1980) cultural typology. The
data instrument developed for the original study (Robbins, & Stylianou, 2001-2002,
Winter) was used to replicate and expand the data in order to determine if any changes
have occurred over the intervening ten years. The instrument included a number of
cultural indicators that were designed to capture the flavor of Hofstede’s five dimensions
in the Web environment. A brief description of each dimension, its characteristics, and
the cultural web site indicators used by the study follows.


Language Dimension

The availability of languages was determined by visiting the home page of each
of the corporations included in this study.

Power Distance Dimension

This dimension relates to the extent to which less powerful members of a society
accept and expect that power is distributed unequally (Hofstede, 1980). The following
characteristics have been adopted from Hofstede (1980):
Low Power Distance
• Distrust of authority figures
• Hierarchy means an inequality of roles, established for convenience
A Longitudinal Study of Cultural, Page 3

---

Journal of International Business and Cultural Studies
• Decision making is consultative in nature
• Powerful people try to look less powerful than they are
High Power Distance
• Acceptance of authority of superiors
• Inequalities among people are both expected and desired
• Less powerful people should be dependent on the more powerful
• Hierarchy means existential inequality between high-ups and lower downs
Clusters with high power distance characteristics include the Latin, Asian, and
Japan and low power distance clusters include the Nordic, Anglo, and German. Cultural
indicators were identified to reflect the characteristics of each of Hofstede’s dimensions.
The presence of organizational charts, biographical sketches of top leaders, and a
corporate message from the CEO were used to characterize the power distance
dimension. It was reasoned that organizational charts remind people of organizational
inequalities (of roles and power) and this is reflected in both the low and high power
distance characteristics. High power characteristics include biographical sketches of top
leaders because these point out the dependence less powerful people have on corporate
leaders, and corporate messages from the CEO because these reflect the acceptance of the
authority of superiors (Robbins, & Stylianou, 2001-2002, Winter).

Uncertainty Avoidance Dimension

Uncertainty avoidance refers to how people respond to uncertainty, ambiguity,
and environmental threats (Hofstede, 1980). Characteristics include:
Weak Uncertainty Avoidance
• Regionalism, internationalism, attempts at integration of minorities
• High labor turnover
• Less ritual behavior
• Few rules as possible
Strong Uncertainty Avoidance
• Nationalism, xenophobia, repression of minorities
• Hierarchical structures of organizations should be clear and respected
• Lower labor turnover less ambitious employees
• More elaborate legal system
Clusters with weak uncertainty avoidance include the Anglo, and Nordic, while
the Latin, German, Asian, and Japan clusters have strong uncertainty avoidance
characteristics. Web site indicators used to represent weak uncertainty avoidance
included: whether the site included a list of job openings; the presence of career/job
opportunity descriptions; and whether the job openings were international in scope.
These measures were chosen because high labor turnover and internationalist attitudes
associate with weak uncertainty avoidance. Indicators for strong uncertainty avoidance
included: presence of a disclosure statement about the use of cookies (the reasoning is
that such statements are more likely in countries with more elaborate legal systems); and
the presence of organizational charts that reflect an organization's hierarchy (Robbins, &
Stylianou, 2001-2002, Winter).

Individualism/Collectivism Dimension
A Longitudinal Study of Cultural, Page 4

---

Journal of International Business and Cultural Studies

This dimension can be defined as people looking after themselves and their
immediate family only versus people belonging to in-groups that look after them in
exchange for loyalty (Hofstede, 1980). Characteristics include:
Individualism
• Everyone has a right to privacy
• Everyone is expected to have a private opinion
• Ideologies of individual freedom
• Greater occupational mobility
• Organizations are not expected to look after employees from the cradle to
the grave
Collectivism
• Press controlled by the state
• Private life is invaded by group(s)
• Private life is invaded by organizations and clans to which one belongs
• Less occupational mobility
• Policies practices based on loyalty and sense of duty

Anglo, Nordic, Latin, and German clusters have individualistic characteristics and
Asian and Japan have collectivistic characteristics. Individualistic indicators include the
need to register to use parts of the site, the capability for secure communication and
transactions providing protection of personal and/or sensitive information; and the
presence of a privacy policy statement (Robbins, & Stylianou, 2001-2002, Winter). These
relate to the idea that everyone has a right to privacy. A collectivistic indicator for this
dimension is the use of cookies, which can be considered an invasion of privacy.

Masculinity/Femininity Dimension

The dominant values in a masculine society are achievement and success whereas
dominant values in a feminine society are caring for others and quality of life (Hofstede,
1980). Characteristics include:
Masculine
•
Men are supposed to be assertive, ambitious, and tough
•
Money and things are important
•
Men should dominate in all settings
•
Dominant values in society are material success and progress
Feminine
• Dominant values in society are caring for others and preservation
• Everybody is supposed to be modest
• Sympathy for the weak
• Conservation of the environment is seen as a more important problem than
economic growth
• Work is less central in people’s lives
Clusters with masculine characteristics include Anglo, German, and Japan.
Clusters with feminine characteristics include Nordic, Latin and Asian. Masculine culture
A Longitudinal Study of Cultural, Page 5

---

Journal of International Business and Cultural Studies
indicators included the availability of the Annual Report on the web site and the inclusion
of financial highlights. These indicators reflect the importance of money and assets and
represent success. Another masculine indicator was the inclusion of a hit/visitor counter
in the web site, an indication of keeping score. Feminine indicators included the
discussion of social responsibility issues and evidence of cultural sensitivity. These
indicators reflect a caring and sympathetic attitude towards others (Robbins, & Stylianou,
2001-2002, Winter).

Short/Long Term Orientation Dimension


This dimension relates to how people view time (Hofstede, 1980).
Characteristics include:
Short-term Orientation
• Respect for traditions
• Concern with ‘face’
• Quick results expected

Long-term Orientation
• Thrift, being sparing with resources
• Large savings , and funds available for investment
• Perseverance towards slow results
Clusters with a short-term orientation include Anglo, Nordic, Latin and German.
Asian and Japan have long-term orientations. Web indicators that reflect short-term
orientation included the use of a site map/index page, and the availability of a search
engine. These factors save time and provide quick results. Other short-term indicators
included the ability to communicate with a corporate expert in a real time on-line session,
the use of a "frequently asked questions" section that provides quick answers, and press
releases. These indicators reflect the desire to have information immediately. Finally, the
presence of a corporate history on the web site is an indication of respect for tradition. A
long-term orientation indicator is the presence of a vision statement on the web site.
Vision statements by definition express the long-term orientation of the corporation
(Robbins, & Stylianou, 2001-2002, Winter).

DATA COLLECTION

A criticism of work done in the area of assessing cultural values relates to the
limited number of countries often included in the studies. Many of the studies only
include two or three countries (Baack, & Singh, 2007; Cho, Kwon, Gentry, Jun, &
Kroop, 1999; Choe, 2004; Gould, Zakaria, & Yusof, 2000; Lin, 2001; Overbeeke, &
Snize, 2005; Singh, & Baack, 2004; Singh, Kumar, & Baack, 2005), which decreases the
ability to generalize the results. To overcome this limitation, ninety global companies
were included in both the 1998 as well as the current study. These organizations were
selected from the list of the 500 largest global corporations published annually by
Fortune Magazine in 1998 and 2007 respectively. This project chose not to attempt to
survey the same set of companies included in the 1998 study but rather collected data
from the current largest global corporations. It was felt that the cultural dimensions
A Longitudinal Study of Cultural, Page 6

---

Journal of International Business and Cultural Studies
found in the web pages would not be unique to the corporation but rather would reflect
the cultural values associated with each cluster assignment. Therefore, it was felt that this
decision would not impact the reliability of the data collected. Fortune Magazine was
chosen because the companies included in their list have successfully faced the
complexity of operating across the globe and have been confronted by cultural issues.
These corporations have had to depend on computer technologies to conduct business and
also have the financial resources to develop complex web sites that reflect and support
the corporate structure.
Hofstede’s (1980) six country clusters were used to categorize companies based
on the location of the corporate headquarters and the top 15 global corporations in each
of the six clusters were included, for a total of ninety (see Table 1). In terms of
cluster/country representation, sixteen countries scattered around the globe were included
in the 1998 study (Robbins, & Stylianou, 2001-2002, Winter) and 20 countries were
included in 2008. The mean revenues for the 90 corporations were $49,909 million in
1998 and $96,682 million in 2008 and the mean number of employees grew from
137,120 and to 209,305 respectively. This represents clear growth both in terms of
manpower and financial resources of global corporations over time.

Table 1: Cluster/Country Representation
Cluster HQ Country
Frequency Percent

‘98
‘08
‘98
‘08
Anglo
U.K.
2
2
2.2
2.2

USA
13
13
14.4 14.4
Nordic
Finland
1
1
1.1
1.1

Netherlands
9
8
10.0 8.9

Norway
1
2
1.1
2.2

Sweden
4
2
4.4
2.2

Denmark
0
2
0
2.2
Latin
Belgium
0
2
0
2.2

Brazil
0
1
0
1.1

France
11
8
12.2 8.9

Italy
3
2
3.3
2.2

Spain
0
2
0
2.2

Venezuela
1
0
1.1
0
German Germany
12
10
13.3 11.1

Switzerland
3
3
3.3
3.3
Asian
China
0
7
0
7.8

India
1
1
1.1
1.1

Korea
12
4
13.3 4.9

Malaysia
1
1
1.1
1.1

Taiwan
1
1
1.1
1.1

Thailand
0
1
0
1.1
Japan
Japan
15
15
16.7 16.7
Total
90
100%


A Longitudinal Study of Cultural, Page 7

---

Journal of International Business and Cultural Studies
The web site of each corporation in the sample was examined to determine if
there is evidence of cultural homogenization over time as a result of conducting business
in a global context. For each corporation in the sample, data was collected responding to
the demographic questions in the survey from public sources, the company’s web site, as
well as Fortune Magazine. The web site of each sample company was then examined to
determine if cultural convergence has occurred since the original study was undertaken.
Table 2 shows the diversity of companies represented in the 1998 and 2008
studies. Twenty industries were represented in the earlier study and twenty-three in the
current.
Table 2: Industry Representation
Industry
Frequency
Percent

‘98
‘08
‘98
‘08
Aerospace
1
1
1.1
1.1
Banks: Commercial and Savings
6
18
6.7
20.0
Chemicals
4
1
4.4
1.1
Computers, Office Equipment
1
1
1.1
1.1
Consumer Food Products
0
1
0
1.1
Diversified Financials
0
1
0
1.1
Electronics, Electrical Equipment
12
9
13.3
10.0
Energy
1
1
1.1
1.1
Food
2
0
2.2
0
Food and Drug Stores
2
4
2.2
4.4
Forest and Paper Products
1
0
1.1
0
General Merchandisers
1
1
1.1
1.1
Insurance
10
0
11.1
1.1
Insurance: Life, Health (Mutual)
0
1
0
1.1
Insurance: Life Health (Stock)
0
5
0
5.6
Insurance P + C (Stock)
0
4
0
4.4
Mail, Package, Freight, Delivery
1
1
1.1
1.1
Metals
1
2
1.1
2.2
Motor vehicles and Parts
14
11
15.5
12.2
Network and other
0
2
0
2.2
Communications Equipment
Petroleum Refining
13
16
14.4
17.8
Pharmaceuticals
1
0
1.1
0
Shipping
0
1
0
1.1
Telecommunications
5
4
5.6
4.4
Tobacco
1
0
1.1
0
Trading
10
1
11.1
1.1
Utilities, Gas and Electric
3
3
3.3
3.3
Wholesalers: Health care
0
1
0
1.1
Total
90
90
100%
100%


A Longitudinal Study of Cultural, Page 8

---

Journal of International Business and Cultural Studies

FINDINGS AND DISCUSSION

Language Dimension

The language of the home page of each of the 90 web pages included in this study
was recorded. The findings indicate that English continues to dominate as the language
of choice (see Table 3, which includes results from both studies). This is in agreement
with what has been reported in other studies (e.g., ( Group, 2008; Lohr, 2000, January 9).
However, in 2008 only 86% percent of the home pages were available in English
compared to 100% in 1998. The few web pages that did not include English on the home
page typically had an English option so that it was possible to collect data from the other
parts of the web site. The number of home pages available only in English has increased,
while the number of those offering multiple language options has declined in 2008.

Table 3: Home Page Language Options

Number of

Home Pages
Percentage

‘98
‘08
‘98
‘08
English Available
90
86
100%
86%
English Only
31
38
34%
42%
Bilingual
54
38
60%
42%
Trilingual
5
7
6%
8%
4 or more
0
7
0%
8%

Changes between 1998 and 2008

Table 4 presents the frequency of occurrence of each Web Site Cultural
Indicator sorted by country cluster. T-tests were run and the shaded cells indicate
significant changes from 1998 to 2008.


Table 4: Frequency Results
Percentage of Occurrence

Dimension
Indicator
Anglo
Nordic
Latin
German
Asian
Japan
‘0
‘9
‘9
‘9

‘98
‘08
‘98
‘08
‘98
‘08
‘08
‘08
8
8
8
8
Organizational
13
7
40
53
40
33
27
27
13
47*
47
73
Chart
Power
Bio-sketches of
73
100
40
80*
67
87
33
33
27
40
7
13
Distance
Top Leaders
*
**
Message from
100
13
40
7
7
20
7
20
27
27
60
47
CEO
**
Uncertainty
Listing of Job
93*
80
87
60
87
20
47
33
20
20
7
27
Avoidance
Openings
*
A Longitudinal Study of Cultural, Page 9

---