Ảnh hưởng của khoảng cách văn hóa lên hiệu quả hoạt động của nhân viên tại các công ty Đa quốc gia

Working Paper 2022.1.2.16

- Vol 1, No 2

ẢNH HƯỞNG CỦA KHOẢNG CÁCH VĂN HÓA LÊN HIỆU QUẢ HOẠT

ĐỘNG CỦA NHÂN VIÊN TẠI CÁC CÔNG TY ĐA QUỐC GIA

Phạm Hồ Nhất Phương¹, Nguyễn Hoàng Anh, Phạm Thị Thanh Lan, Võ Anh Nhựt Minh,

Đinh Thị Thư

Sinh viên K57A - Quản trị kinh doanh quốc tế

Cơ sở II Trường Đại học Ngoại thương tại TP. Hồ Chí Minh, Việt Nam

Lê Sơn Đại

Giảng viên Bộ môn QTKD - TCKT

Cơ sở II Trường Đại học Ngoại thương tại TP. Hồ Chí Minh, Việt Nam

Tóm tắt

Nền kinh tế Việt Nam đang phát triển nhanh chóng và thu hút một lượng lớn sự quan tâm của các

nhà đầu tư nước ngoài. Tuy nhiên, hiệu quả lao động ở Việt Nam vẫn là một rào cản đối với các

nguồn vốn từ nước khác. Qua nghiên cứu của này, mối quan hệ tiêu cực giữa khoảng cách văn hóa

và hiệu quả làm việc của nhân viên tại các công ty đa quốc gia đã được chứng minh. Biến ngành

công nghiệp là một biến kiểm soát đối với hiệu quả làm việc của nhân viên. Đồng thời biến khoảng

cách thể chế điều tiết tiêu cực đối với mối quan hệ này. Từ kết quả nghiên cứu, các khuyến nghị

được đưa ra cho các công ty đa quốc gia bao gồm nên chú ý lựa chọn địa điểm để mở rộng quy mô,

nên ưu tiên những quốc gia có văn hóa tương đồng với công ty mẹ để giảm khoảng cách văn hóa.

Bên cạnh đó, nhà quản lý có thể giảm bớt tác động của khoảng cách văn hóa bằng việc lựa chọn

phương thức thâm nhập phù hợp, các phương thức thâm nhập ít sự kiểm soát và ít rủi ro thấp như

cấp phép, liên doanh, hay mua lại và sáp nhập sẽ phù hợp hơn khi tham gia vào các thị trường có

khoảng cách văn hóa cao.

Từ khóa: Công ty đa quốc gia, kinh doanh quốc tế, hiệu quả hoạt động, nhân viên

IMPACTS OF CULTURAL DISTANCE ON

PERFORMANCE OF EMPLOYEES IN MNCs’ SUBSIDIARIES

Abstract

Performance of employees is described as the “ultimate dependent variable” in human resource

management, enhancing it is the priority of corporations to create long-term competitive

advantages. Furthermore, recent literature stated that cultural distance is an important determinant

of employee performance. This paper aims to clarify the impacts of the cultural distance between

home and host country on the performance of employees in multinational corporations’

subsidiaries in Vietnamese market. This study also investigates the moderation effect of career

¹Tác giả liên hệ, Email: [email protected]

FTU Working Paper Series, Vol. 1 No. 2 (2/2022) | 223

experience and institutional distance on the mentioned relationship. Our results support the

hypothesis that cultural distance has a significant negative impact on employee performance and

this relationship is affected by the industry sector. In addition, the relationship is also moderated

by experience and institutional distance. These findings will contribute in helping MNCs improve

their employee performance by mitigating the impact of cultural distance through choosing a

suitable country and entrance modes to expand.

Keywords: Cultural distance, employee performance, MNCs, institutional distance

1. Introduction

Culture has long been proven as a barrier to the internationalization process of multinational

corporations (MNCs) in the world (Beugelsdijk, 2018). Companies have to deal with several

critical decisions such as entry modes (Tihanyi et al., 2005), international diversification (Grosse

and Trevino, 1996), subsidiary management (Roth and O’Donnell, 1996), and human resources

management (Morosini et al., 1998). Failure to understand elements surrounding the subsidiary

national culture would lead to an adaptation crisis (De Mooij & Hofstede, 2010). In addition, in

order to evaluate the efficiency of a subsidiary, performance of employees is among the most

essential aspects when operating business (Brayfield & Crockett, 1955). As a result, more and

more scholars have been drawn attention to the relationship between employee performance and

cultural distance (Shenkar, 2001; Froese & Peltokorpi, 2011; Ahammad et al., 2016).

Additionally, in previous studies, the main concern of the authors was the relationship

between cultural distance and job performance (Morosini, 1998; Froese & Peltokorpi, 2011).

However, the effects of control and moderate variables have not been mentioned (Dikova & Sahib,

2013; Shahzadi et al., 2014; Meyerson & Dewettinck, 2012) in these researches. Therefore, this

research aims to observe these variables: whether they have an effect on the relationship and how

the impact is as well as whether there are other factors moderating this relationship such as

industrial, direct manager's nationality, etc. Because there are studies only mention a relationship

but not take into account the external environment, including many factors that can moderate this

relationship, which results in impurity in research (Chen et al., 2010). If the result supports our

hypothesis, it will be an important contribution for MNCs in Vietnam to find solutions to optimize

employee’s productivity.

To bridge the mentioned gaps above, this research aims to discover the direct relationship

between cultural distance and performance of employees on subsidiaries of multinational

corporations. Consequently, this paper gives a firm positive statement on the relation between the

two variables. Finally, based on the results, MNCs can take into account the suggested implications

when expanding to other territories.

This study uses the standard framework of Perry (1998) which has 5 sections widely applied

in academic literature. In detail, section 1 identifies the elements that make the topic worth

researching and deliver contexts as well as rationale for the research. Section 2 provides review of

current literature on the contexts: Employee performance and cultural distance. From that, a

theoretical model with hypotheses was proposed to bridge the gaps of those literatures. It is

followed by section 3 which justifies the methodology and the data analysis process used to

examine the proposed theoretical model. A report of the results is presented and key findings are

also highlighted in section 4. Eventually, section 5 consists of the conclusion with a brief summary

FTU Working Paper Series, Vol. 1 No. 2 (2/2022) | 224

of the paper. It also proposes implications for MNCs when expanding to other countries and

discusses limitations as well as suggestions for further research.

2. Literature review

Kogut and Singh are the firsts to lay the foundation of cultural distance when they mention

the differences in entry modes because of the influence of culture factors (Kogut and Singh, 1988).

It was described to be the extent to which the cultural norms in one country have a significant

difference when compared to another country (Kogut and Signh, 1988). Using Hofstede’s

multidimensional culture framework, Kogut and Singh (1988) introduced a Euclidean distance

measure to capture cross-country cultural differences in one index. The Euclidean distance index

takes the difference on the national score on each of Hofstede’s cultural dimensions (Hofstede,

1980) and then aggregates these differences in one overall index. Cultural distance is calculated as

the distance to a single country. The vast majority of cultural-distance studies follow this approach

in operationalizing and measuring cultural distance (Kirkman et al., 2006; Kirkman, Lowe, &

Gibson, 2017).

Culture governs the way of thinking and the soul of humans (Hofstede, 2010). Hofstede

identiﬁed six major dimensions to explain differences among cultures. These are: (1) Power

distance; (2) Individualism and collectivism; (3) Masculinity; (4) Uncertainty avoidance; (5)

Long-term orientation; (6) Indulgence.

Employee performance is the process of an individual using his/her own skills, experience,

knowledge, personality and abilities (Vroom, 1964; Hunter, 1986) with a view to reach their goals.

In other words, Azar & Shafighi (2013) described that job performance is a set of behavioural and

functional patterns involving knowledge, skills, managerial competence, conscience and cognitive

abilities in the work environment - the same words of Rothmann et al. in 2002. Employee

performance has been also described by Motowidlo (2003) as the overall values accumulated by

an individual during a specified period of time as discrete pieces of behaviours expected by the

organization. Employee’s performance is the combining of quality and quantity of the work or

task that employees executed to fulfill a specific target (Shahzadi et al, 2014).

The final strand examines the influence of cultural distance on subsidiary

performance. Accordingly, the major studies indicate that due to the complexity and uncertainty

of doing business in a distant host country, cultural distance is expected to have a negative effect

on employee performance. This perspective is proved through the studies of Maseland et al. 2018;

Beugelsdijk et al. 2018; Brouthers, Marshall & Keig, 2016; Harzing & Pudelko, 2016; van Hoorn

& Maseland, 2016). In addition, there are a number of factors studied that have an impact on

employee performance and are influenced by cultural distance as well, such as organizational

culture (Kawiana, et.al., 2018; Szydło & Grześ-Bukłaho, J., 2020), cross-cultural training

(Shahzadi et al., 2014; Kempf, & Holtbrügge, 2020), interests and reward management system

(Lent et al., 1994; Sullivan & Hansen, 2004; Mehmood et. al., 2013) and so on.

Based on previous literature, this research predicts a negative relationship between cultural

distance and employee performance in MNCs. With the different approaches mentioned above and

with gap not being studied, we hypothesize that:

H1: Cultural distance between home-host countries has a significant negative impact on

employee performance of MNCs’ subsidiaries.

FTU Working Paper Series, Vol. 1 No. 2 (2/2022) | 225

There is a large number of studies on employee age and performance, and they seemed to fail

to identify a clear relationship between them. While some research showed a positive relationship

between the performance of employees and their age, the others showed a negative one (Ng and

Feldman, 2008). Waldman and Avolio (1986), in their meta-analysis with 40 samples in 13

empirical studies, found that age has a positive impact on employee performance at productivity

measures but a negative impact on employee performance at supervisor ratings. Thus, we also

consider age is one control variable and it is necessary to delete its impacts from employee

performance.

According to Wang and Giouvris (2020), each different industry has its own characteristics.

And foreign firms belonging to different industries (manufacturing vs non-manufacturing) are

differently affected by cultural distance and country risk (Wang and Giouvris, 2020). Specifically,

according to Drogendijk and Slangen (2006), the impact of cultural distance and country risk on

manufacturing companies is less than non-manufacturing companies. Through this, we come to a

conclusion that each industry has its own characteristics and this affects the impact of the

independent variables (in this paper, cultural distance).

A number researchers suggest explanations for these findings that women have more

sensitiveness and interests in developing relationships with colleagues than their man counterparts

(Major and Adams, 1983; Major et al., 1984; Rosener, 1990) or men have a higher level of

participation and performance along with the competitiveness than women do (Croson and

Gneezy, 2009). These gender differences in abilities, skills, style of work, orientation, participation

make each gender employee performance different (Croson and Gneezy, 2009), so that we

consider it is also one control variables which we need to remove to purify the relationship between

to our independent variables (cultural distance) and employees performance

For the above reasons, the hypothesises are formulated:

H2: The relationship between cultural distance and MNCs employee performance is

significantly controlled by industry.

H3: The relationship between cultural distance and MNCs employee performance is

significantly controlled by gender.

H4: The relationship between cultural distance and MNCs employee performance is

significantly controlled by age.

Several theories have laid the foundation of perception about the relationship between job

experience and performance. If employees can exploit themselves in experience, they can improve

their ability, and in progress influence employees’ job performance (employee performance)

(Ehrenberg & Smith, 2000). Employee experience is also predicted to be able to enhance

performance of employees (Weiss & Bloom, 1990). Both opinions have the same idea that

employee performance changes overtime because individuals accumulate job experience.

In this paper, we will only mention the first aspect of institutional distance, scholars have

argued that focussing on all three pillars of institutional distance rather provides a broad basis for

analysis, thus leading to oversimplification (Zaheer et al., 2012). Differences in the regulative

pillar of institution between the home and host countries have influence on how economic actors

interact (North, 1990). One of the cultural orientations that helps explain a key difference between

the United States and Germany, for example, is the attitude towards uncertainty. In Hofstede’s

FTU Working Paper Series, Vol. 1 No. 2 (2/2022) | 226

study (1980) Germany has an uncertainty avoidance index of 67, whereas the United States has an

uncertainty avoidance index of 43. That means Americans are more comfortable with uncertainty

than Germans such as they will have fewer laws in comparison with Germans wanting certainty

and protection from risk. Hence, we hypothesize:

H5: Job experience significantly negatively moderates the relationship between culture

distance and employee performance.

H6: Institutional distance significantly negatively moderates the relationship between culture

distance and employee performance.

Figure 1. Conceptual model

Source: Authors

3. Methodology

To measure the employee performance, this research uses surveys as the main source to collect

primary data from employees of MNCs’ subsidiaries in Viet Nam. Two parts are included in the

instrument survey. Section A covers the demographic of both the respondent and the company

where he/she works. The objective of this part was to collect the control variables data and

demographic information. Section B is a range of criteria to assess employee performance

(employee performance), these criteria are based on the past literature. There are 6 criteria with

sub-criteria written in statements. These statements will be graded through a 5-point Likert scale.

These points in the scale are “Strongly Disagree” (1), “Disagree” (2), “Uncertain” (3), “Agree”

(4), “Strongly Agree” (5).

Table 3. Sources of questionnaire

No. Items

Content

References

EP1 Task evaluation

I was able to schedule my job so that it Kiker

&

was finished on time.

Motowidlo, 1999

I adequately performed all the tasks

assigned

I carried out duties that are required of

me

FTU Working Paper Series, Vol. 1 No. 2 (2/2022) | 227

No. Items

Content

References

I managed to distinguish between main

issues and side issues

My work is done with an optimal

efficiency

I engage in activities that will directly

affect my performance’s evaluation

My attendance at work is above the

norm

EP2 Practice

to

get I assumed additional duties.

Arvey and Mussio

(1973)

higher result

I do not wait for instruction to start my

next tasks

When available, I took on demanding

work duties

I actively involved in challenging

projects/tasks at work

I performed extra duties at work

EP3 Innovate to get I've been trying to keep my job skills Feirong Yuan and

higher result

and knowledge up-to-date

Richard

W.

Woodman (2010)

I contributed creative solutions to fresh

issues

I have been constantly searching for

new ways to enhance my results.

I learned how to easily overcome tough

circumstances and setbacks.

EP4 Team’s

performance

I passed along information to coworkers Lecturers, Institute

effectively

of

Management

Sciences, Kohat

I actively help other team members with

heavy workload

University

Sciences

of

&

I made effort to contribute in the

company’s result

Technology (2011)

I satisfy in my team’s result

I confidently persuade and defend my

idea

FTU Working Paper Series, Vol. 1 No. 2 (2/2022) | 228

No. Items

Content

References

I teamwork with other colleagues very

well

EP5 Counterproductive

work behavior

I grumbled at minor problem at work

I exaggerated the problem at work

Koopmans et al.

(2014)

I should have paid more attention to the

positive aspects of a work situation

I talked about the negative aspects of

my job with colleagues.

I didn’t complete the assigned task

I am able to refuse doing tasks easily

EP6 Compare

colleague

with I receive a higher bonus/commission Bishop (1987)

due to exceeding kpi than other

colleagues with the same position.

I work more efficiently than most of my

colleague

My performance is better than that of

my

colleagues

with

similar

qualifications.

I receive promotion faster than those

colleagues who entered the company at

the same time with me

My overtime hours is higher than the

average of that of my coworkers

I have better skills in communication

and socialization than other colleagues

in my department

The current research adopts six dimensions of national culture measurement which is based

on extensive research done by Professor Geert Hofstede, Gert Jan Hofstede, Michael Minkov and

their research teams (2006). In order to calculate the distance in terms of each cultural dimension

between the parent firm and its subsidiary in Vietnam, we use the absolute value of the result taken

from indicators of Vietnam that subtracts the equivalent cultural dimension indicators of the host

country (so that a positive real number of distances will be collected). We shall take the data of

each cultural dimension index on the scale of 100 in each country on Hofstede Insight. Before

analyzing data, we also normalise all results of distance calculated above with a logarithm

(Quackenbush, 2002).

FTU Working Paper Series, Vol. 1 No. 2 (2/2022) | 229

After completing the description analysis, Cronbach Alpha is applied to judge the reliability

of a measurement which is set to the internal consistency of this measurement (Hair et al, 2016),

all of the variables were gone into exploratory factor analysis (EFA) to investigate the information

and propose the number of components expected to represent the data (Hair, Black, Babin and

Anderson, 2014). Afterward, Confirm factor investigation (CFA) was implemented to assess a

confirmatory test of appraisal (Saunders, Lewis, and Thornhill, 2016), which guarantees that the

deliberate factors are convincing and consistent to represent constructs related to the hypothetical

model (MacCallum and Austin, 2000). The next step is structural equation modeling (SEM)

technique to determine the extent of fit of the entire model consisting of all the constructs (Hair et

al, 2006). Followingly, bootstrap embedded methods are carried out to define the model fit. After

that, multigroup analysis (MGA) is conducted to evaluate whether two or more variables have the

same relation in groups. Finally, we test the effect of the control and moderating variables for

justifying hypotheses H2, H3, H4 and H5.

4. Results

4.1. Sample profile

Our sample profile classification is based on the country of the parent company (which have

subsidiaries in Vietnam), industry in which employees work, gender and age of employees, and

their experience in the MNCs subsidiary also. The detailed sample descriptions are present in table

1 below.

Table 1. Sample profile

Dimensions

Home country Number of home country

Highest percentage home country

Number of observation Percent (%)

21

United States

30.28

11.5

19.7

17.0

51.8

39.4

36.2

24.3

42.2

57.8

72.5

23.4

4.1

Industry

Primary

25

43

37

113

86

79

53

92

126

158

51

9

Secondary

Tertiary

Quaternary

From 18 to 24

From 25 to 34

From 35 to 44

Male

Age

Gender

Female

Experience

Less than 1 year

From 1 to 5 years

From 5 to 10 years

Table 2. Descriptive Statistics

FTU Working Paper Series, Vol. 1 No. 2 (2/2022) | 230

Minimum Maximum Mean Std. Deviation

Employee performance

Task evaluation

1.04

1.03

1.16

.98

4.93

4.91

5.00

2.7948 .85954

2.8842 .90210

2.6983 .79482

2.9964 .87270

2.8107 .79882

2.8786 .83180

Practice to get higher result

Innovate to get higher result

Team’s performance

Counterproductive work behavior

Compare with colleague

Cultural Distance

1.03

Power distance

.05

.00

.05

.00

3.55

1.2993 .75829

1.5482 1.08592

1.4234 .96451

1.6089 1.14650

1.5089 1.05717

1.3998 .88705

Individualism distance

Masculinity distance

Uncertainty avoidance distance

Long term orientation distance

Indulgence distance

Institutional Distance

Total observations

.59

4.36

2.2608 .76021

218

On the hand, about cultural distance among observations in this paper, the smallest one is 0

and the highest one is 3.55 in Likert scale 5-point and 71 in 0-100 scale. Identically, institutional

distance respectively is 11.8 and 87.2 in 0-100 scale or 0.59 and 4.36 in Likert scale 5-point.

Details and statistical description are presented in table 1 and table 2.

4.2. Testing measurement model: Confirm factor analysis (CFA) result

The following step after descriptive statistics analyses is accessing Cronbach Alpha. As

introduced in section 3, Cronbach Alpha is commonly applied to assess the reliability of a

measurement which determines the internal consistency of this measurement. According to Hair

et al. (2016), Cronbach Alpha needs to be above 0.7 and low item total correlations (less than 0.3)

and low factor loadings (less than 0.4) must be removed. Applying this model, we can conclude

that the measurement constructs meet the reliability requirement as all construct measurements

had Cronbach Alpha (0.903) above 0.7 and item total correlations above 0.3.

Next, with the aim of exploring the data and providing information about how many factors

are needed to best represent the data based on the statistical method, exploratory factor analysis

(EFA) with Maximum Likelihood Method is carried out. The Kasier Meyer-Olkin Measure of

FTU Working Paper Series, Vol. 1 No. 2 (2/2022) | 231

Sampling Adequacy (KMO) and Bartlett’s Test of Sphericity (BTS) were assessed to ensure the

appropriateness of the data for EFA. The KMO coefficient was higher than 0.80 and BTS was

significant at the 0.05 level, indicating the adequacy of the items (Hair, Black, Babin &

Anderson, 2010). The break in the scree plot and extracted eigenvalues (greater than 1) suggested

a six-factor solution. The items with factor loadings lower than 0.50 or cross-loaded items were

removed one at a time and remaining items were factor analyzed again (Hair et al., 2010).

In this model, the Kaiser-Meyer-Olkin (KMO) index is 0.871 (above 0.5) and Bartlett’s test

has p<0.001 (smaller than 0.05). The total variance elaborated was approximately 60.5% which

exceeded the threshold value of 60% and thus suggested that the total validity of the scales was

reasonable (Hu, 1999). The EFA results satisfied the requirement for the reliability coefficient of

the measurement scales, indicating high internal consistency (Hair et al., 2010). Subsequently, the

factors suggested from the EFA based on empirical data and the items associated with each factor

are similar to that of the proposed measurement model for CFA.

Table 3. KMO and Bartlett’s Test

Kaiser-Meyer-Olkin Measure of Sampling Adequacy

.871

Bartlett's Test of Sphericity Approx. Chi-Square

1609.039

66

df

Sig

.000

As suggested at section 3, CFA is accessed with a view to providing a confirmatory test of

the measurement theory (Hair et al. 2014). In reality, CFA is used to test how well the measured

variables represent the pre-specified constructs (variables load on specific constructs). While

EFA enables the statistical method to determine the number of factors and loadings, CFA

articulates how well our conceptual specification of the factors matches with the actual data. In

other words, researchers can apply CFA to confirm to reject the predetermined theory (Hair et

al. 2014).

Table 4. Total Variance Explained

Total Variance Explained

Initial Eigenvalues

Extraction Sums of Squared Loadings

Component Total % of Variance Cumulative % Total % of Variance Cumulative %

1

2

3

4

5

6

3.998 66.637

.623 10.381

.428 7.137

.387 6.456

.339 5.648

.224 3.741

66.637

77.018

84.155

90.611

96.259

100.000

3.998 66.637

66.637

Extraction Method: Principal Component Analysis.

FTU Working Paper Series, Vol. 1 No. 2 (2/2022) | 232

Specifically, the test for single constructs (Inno, Counter, Prac, Task, Team, Compare)

performed a significant fit to the data with all model fit index meeting the requirement N=218,

2/df<2, p<0.05, GFI>0.9, CFI>0.9, and RMSEA<0.8, (Hair et al, 2010). However the Goodness

of fit is designed for testing the entire model, not an individual construct at a time, therefore it is

poor practice to evaluate the measurement model fit through a separate analysis for each construct

instead of one analysis for the entire model (Hair et al, 2014).

Figure 3. Initial CFA

(χ²= 187.480; χ²/df = 3.537; DF = 53; N=218; p<0.001; GFI = 0.879; CFI = 0.915 and

RMSEA = 0.108)

The model fit index told that the measurement model fit did not fit the data well. The 2/df and

GFI indicators do not meet the requirement of Hair et al. (2010).

Therefore, model diagnostics, namely path estimates, standardized residuals, and modification

indices were employed to improve the model fit (Hair et al. 2006). The final result is reported in

the figure below:

Figure 4. Final CFA

FTU Working Paper Series, Vol. 1 No. 2 (2/2022) | 233

(χ²= 81.022; χ²/df = 1.929; DF = 42; N=218; p<0.001; GFI = 0.943; CFI = 0.975 and

RMSEA = 0.114).

After modification, both GFI and CFI exceeded the 0. thresholds while 2/df, p, RMSEA are

all under the thresholds required by Hair et al, 2010.

In conclusion, these results above of the model fit and construct validity present an appropriate

measurement model (which meets all the requirements). The measurement model, therefore, is

acceptable for hypothesis testing step.

4.3. Hypotheses testing results

The following figure illustrates the SEM result:

Figure 5. SEM

According to the CFA outcome, statistic of maximum likelihood estimation (MLE) of

estimating the parameters of a probability distribution by maximizing a likelihood function

(Myung, I. J., 2003), so that under and specify the thesis structural model. Consequently, these

control variables: Gender (female or male), experience (work longevity) and industry (a dummy

variable) are demonstrated. In overall, not only does the index meet the acceptable benchmark

(according to Bauldry, 2015 and Brett, 2010, acceptable benchmark should be above 0.8) but it

also qualifies for a well fit model (above 9.0) (Hair et al, 2014): CFI=0.975, GFI=0.943.

Table 5. Regression Weights of the sample

Dependent variables

Estimate S.E. C.R.

p

Result

EP

<---

CD

EP

CD

-,326

1,084

,938

,777

,970

,748

,053 -6,205 *** Supported

,077 14,032 ***

EP5

EP2

EP1

EP4

CD3

,083 11,316 ***

,079 9,842 ***

,059 16,401 ***

,072 10,458 ***

FTU Working Paper Series, Vol. 1 No. 2 (2/2022) | 234

Dependent variables

Estimate S.E. C.R.

p

Result

CD4

CD1

CD2

EP3

EP6

CD5

CD6

<---

CD

EP

,763

1,085

,803

,066 11,499 ***

,086 12,689 ***

,138 5,808 ***

1,000

1,093

1,000

,605

EP

,079 13,903 ***

CD

,063 9,670 ***

Table 6. Standardized Regression Weights and Squared Multiple Correlations

Dependent

variables

Estimate

Variables

Estimate

EP

<--- CD

<--- EP

-,432

,897

EP

,187

,357

EP5

CD6

EP2

EP1

EP4

CD3

CD4

CD1

CD2

EP3

EP6

CD5

CD6

<--- EP

<--- CD

<--- EP

<--- CD

,688

,617

,741

,679

,875

,835

,608

,832

,905

,807

,597

CD1

CD3

CD4

CD5

CD2

EP1

EP6

EP2

EP5

EP4

EP3

,766

,460

,549

,819

,652

,697

,381

,473

,805

,369

,692

Within the model, the negative impacts of cultural distance on employee performance in

MNC’s subsidiary have been confirmed. To be specific, the Regression Weights index proves that

there is a meaningful relationship between cultural distance and employee performance (p<0.001).

Moreover, the estimates index of this table also demonstrates the negative relationship (<0). As

such, hypotheses H1 (Cultural distance between home-host country has a significant negative

impact on employee performance of MNC’s subsidiary) is accepted.

We use the Standardized Regression Weights table to assess the impact level of the

independent variables (cultural distance) on the dependent variable (employee performance). The

table indicates that by increasing or decreasing one standard deviation in the cultural distance

FTU Working Paper Series, Vol. 1 No. 2 (2/2022) | 235

variables, the employee performance would react in the opposite direction with a 0,432 standard

deviation change. The R-Squared index (R²= 0,187) shows that 18.7% of the variance of the

employee performance is explained by the cultural distance.

4.4. Robustness check (Moderation analysis)

Robustness check is used to define the model fit and there are many methods to conduct it. In

this paper, we test the robustness of the model by using bootstrap method. The result of Bias/SE-

Bias is 1 and it is less than 1.96, which means the using model is optimal to represent the

relationship of cultural distance and EP. Details are represented in Table 7.

Table 7. Bootstrap results

Parameter

EP

SE

SE-SE

.001

.002

.000

.001

.000

.001

Mean

-.325

1.089

.939

Bias

.001

.005

.000

.002

.003

.000

.007

.000

.002

SE-Bias

.001

.002

.001

.002

.001

.002

.003

.000

.001

.000

.001

<--- CD

<--- EP

<--- CD

<--- EP

<--- CD

.062

.091

.080

.076

.064

.084

.077

.090

.157

.000

.075

.000

.066

Counterproductive

Practicet

Compare

.778

Uncertainty

Masculinity

PD

.972

.751

.766

Taskevaluation

Team

1.088

.805

Innovate

1.000

1.100

1.000

.607

Longterm

Individual

Indulgence

4.5. Multigroup covariances

Table 8. Control effect of gender, age and industry on employee performance in subsidiaries

results

Control effect of gender

Control effect of

age

Control effect of

industry

Value

Chi-square

df

Chi-square

df

Chi-square

df

171

Constrained

model

151,461

85

228,340

128

232,436

Free model

151,461

84

225,299

126

221,955

168

FTU Working Paper Series, Vol. 1 No. 2 (2/2022) | 236

Difference

P-value

1,067

1

3,041

2

10,481

3

0,3016

0,2186

0,01489

With regard to control variables, a multigroup structural equation modeling approach was

used to compare gender, industry and age on the relationship of cultural distance and employee

performance. The empirical results revealed that correlation between gender (p=0.3016>0.05), age

(p=0.02186>0.05) and employee performance appears to be not supported (p=0.858>0.05).

Therefore, this study does not accept H3 (the relationship between Cultural distance and MNCs

Employee performance is significantly controlled by gender) and H4 (the relationship between

Cultural distance and MNCs Employee performance is significantly controlled by age).

On the other hand, p-value of industry as a moderating variable is 0.034 (p<0.05). Hence, this

study chooses the free model to read the results because of the higher compatibility. In conclusion,

industry has a positive influence on performance of employees in MNC’s subsidiary which support

hypothesis H2 (the relationship between Cultural distance and MNCs Employee performance is

significantly controlled by industry). In detail, the difference when using industry as a control

variable is shown and interpreted below.

Figure 6. Free model of the relationship between Cultural distance and MNCs Employee

performance with respect to industry as a control variable

Table 9. Multigroup model result when control variable of industry is added

Industry 1

Industry 2

Industry 3

Industry 4

SRW

Dependent

variables

SRW

p

value

-,651

,980

value

-,389

,946

value

-,620

,850

value

-,358

,922

EP ← CD

EP5 ← EP

EP2 ← EP

EP1 ← EP

,006

***

,026

***

,779

,480

,002

,872

,651

,582

***

,460

,002

,779

***

,605

***

FTU Working Paper Series, Vol. 1 No. 2 (2/2022) | 237

Industry 1

SRW

Industry 2

SRW

Industry 3

SRW

Industry 4

SRW

Dependent

variables

p

value

,774

,440

,912

,777

,516

,889

,776

,670

,409

value

,708

,807

,861

,876

,418

,667

,939

,757

,696

value

,669

,844

1,014

,812

,896

,995

,838

,793

,703

value

,762

,614

,830

,909

,524

,793

,941

,814

,580

EP4 ← EP

CD3 ← CD

CD4 ← CD

CD1 ← CD

CD2 ← CD

EP3 ← EP

EP6 ← EP

CD5 ← CD

CD6 ← CD

***

,058

***

,006

***

,092

***

,005

,003

,050

***

Squared multiple

Correlation (R2)

,423

,151

,385

,128

According to the multigroup structural equation modeling results, industry operates primary

activities (p=0,06>0.001) and secondary activities (p=0,026>0.001) seems not subjected to the

impact of cultural distance. It can be explained that primary industry which includes agriculture,

forestry; fishing and mining (Kelton et al., 2008; Dalziel, 2007) has the characteristic of heavy

workload with very little need of creativity to fulfil the task (Yletyinen et al., 2019). Employees in

this type of industry need to strictly follow the procedure so that they could harvest the best cereal

crop or catch the biggest fish in the sea, ect. (Tukel & Rom, 1998). This leads to a standardized

procedure applying for everyone in the corporation. Eventually, when everyone knows what they

need to do thoroughly. Their performances are no longer affected by cultural distance. The same

reason can be referred to secondary industry (construction and manufacturing activities), where

standardization is the core key to operate effectively. Another reason may be derived from the

small amount of respondents collected. Lacking of sufficient samples can lead to failure in getting

the real insights through the data (Morse, 2000).

As for industries operating tertiary activities and quaternary activities, employee performance

in subsidiaries of the two industries depends heavily on cultural distance. The R-squared index of

the correlation in industry type 3 is higher than the constraint model, R²=0,385 which means that

38.5% changes in the variance of the employee performance is explained by the cultural distance

in the tertiary industry. Quaternary industry saw a smaller R-squared index compared to industry

type 3. 12.8% of changes in variance of employee performance is explained by the cultural distance

(R²=0,128) in industry type 4.

4.6. Moderation analysis

FTU Working Paper Series, Vol. 1 No. 2 (2/2022) | 238

Table 10. Interaction between cultural distance and employee performance when adding

institutional distance to the model

Relationship Estimate S.E.

ZEP ← ZCDxID -.200

C.R.

p

.062 -3.225 .001

.118 2.853 .004

.076 13.277 ***

ZEP ← ZCD

ZEP ← ZID

.337

1.012

Table 11 & 12. Standardized Regression Weights and Squared Multiple Correlations when adding

institutional distance to the model

Standardized Regression WeightsEstimate Squared Multiple CorrelationsEstimate

ZEP

← ZCDxID

← ZCD

-.351

.337

ZCDxID

ZCD

.000

← ZID

1.012

ZCDxID

ZEP

.000

.680

To test the hypothesis that clarify impact of cultural distance on employee performance and

more specifically whether work experience and institutional distance moderate the above

relationship, a hierarchical multiple regression analysis was conducted. When the interaction term

between institutional distance was added to the regression model, which accounted for a significant

proportion of the variance in employee performance evaluation, ΔR2 = 0,187, p = 0.001 (p<0.05),

b = -0.351. Examination of the interaction plot showed a diminishing effect that as institutional

distance between the home and host country increased, their performance decreased in that

subsidiary. The higher the institutional distance presented, the lower effectiveness of the employee

performance and vice versa. Therefore, the results of interaction analysis of moderate effects of

institutional distance support hypothesis H6.

Table 13. Interaction between cultural distance and employee performance when adding work

experience to the model

Relationship

Estimate S.E. C.R.

p

ZEP ← ZEXP

.174

.141 1.235

.219 -.714

.217

.475

ZEP ← ZCDxEXP -.157

ZEP ← ZEXP -.254

.166 -1.527 .127

In the meanwhile, results show that there is no significant effect of work experience in

moderating the main relationship. When the interaction term between work experience was added

to the regression model, the main relation remained substance. Examination of the interaction plot

FTU Working Paper Series, Vol. 1 No. 2 (2/2022) | 239

showed an enhancing or decreasing effect of work experience (p=0.732>0.05) shows no changes

in the relation of cultural distance and employee performance (p = 0,0217, p > 0,05). Therefore,

the results of interaction analysis of moderate effects of institutional distance and work experience

in hypotheses H5 are not accepted. However, by not changing the impact of cultural distance on

employee performance when adding these elements into the regression model, the result

strengthens the relationship between cultural distance and employee performance.

5. Conclusion

5.1. Result summary

The existing literature of MNC internationalization recognizes the strategic importance of

cultural distance in employee performance; however the topic is still under debate. Acknowledge

the gap, this dissertation is carried out to answer the question “What is the impact of cultural

distance between home and host country on performance of employees in MNC’s subsidiaries?”.

In short, to investigate the relationships between cultural distance of host and home country

with employee’s performance, an in-depth literature review is conducted. Accordingly, the

conceptual model is established with three hypotheses below:

H1: Cultural distance between home-host countries has a significant positive impact on

employee performance of MNCs’ subsidiaries.

H1: Cultural distance between home-host countries has a significant negative impact on

employee performance of MNCs’ subsidiaries.

H2: The relationship between Cultural distance and MNCs Employee performance is

significantly controlled by Industry

H6: Institutional distance significantly negatively moderates the relationship between culture

distance and employee performance

Overall the result indicates that the hypothesis mentioned above is acceptable. In other words,

cultural distance has remarkable impacts on performance of employees in MNC. Particularly,

cultural distance affects employee’s performance in MNC’s subsidiaries in a negative trend.

Furthermore, Industry is a significant control variable in this relationship. And institutional

distance moderates negatively to the relationship between cultural distance and employee

performance. This result is consistent with the dominant prior research related to these contexts.

5.2. Theoretical contributions

Firstly, all previous studies have been done in leading developed countries (mostly are the

United States, Japan and the European Union where multinational companies are headquartered

according to Kim & Park, 2014) and subsidiaries of MNCs there. This paper is conducted in the

context of subsidiaries in Vietnamese, a developing country, which contribute to the study of CDs

and their effects to be more comprehensive.

Secondly, there is a fact that cultural distance's impact is positive or negative on the global

operations of multinationals is still under arguments. Some believe that cultural distance has a

negative impact on most of these activities, others believe that these effects are positive, while the

rest think that cultural distance has no impact at all or that these effects are declined (Stöttinger

and Schlegelmilch, 1998; Mitra and Golder, 2002; Sethi et al., 2003). By affirming the negative

FTU Working Paper Series, Vol. 1 No. 2 (2/2022) | 240

impact of cultural distance on employee performane, this study contributes to give more points

and evidence to resolve the arguments above.

Thirdly, studying the factors affecting employee performane but not eliminating the effects of

the control variable may be the reason for the inconsistent results of these studies (Dikova & Sahib,

2013; Shahzadi et. al., 2014; Meyerson, & Dewettinck, 2012). Accordingly, this paper uses the

multigroups method to remove these noises from the measurement results, making conclusions

more accurate, then contributing to explain the current inconsistent results.

Finally, by incorporating moderating variables and controlled variables in one model, this

paper contributes to the development of these quantitative methods which are commonly used (Wu

& Zumbo, 2007). The incorporation of moderators and controllers in the model also contributes to

clarify the relationship of cultural distance and employee performane as well as the effects of other

factors on this relationship.

5.3. Implications

From the paper and literature review, the following section provides some implications for

management in multinational corporations. As a result of the current research sample, the

suggestion ocuses on how to mitigate the impact of cultural distance so that the employee can

work more efficiently.

Overall, MNCs should pay attention to cultural distance, especially MNCs that operate in the

tertiary and quaternary industry, to achieve the best outcome when expanding to a new

international market. Firstly, cultural distance has a significant negative impact on employee

performance; therefore, worsening the outcome of organization. It is advised that when choosing

a location to operate a new subsidiary, the priority should be placed on those countries that have

similar culture to the headquarters to reduce the cultural distance. However, after considering all

advantages and disadvantages of entering a new market among other available options, under the

circumstances of having no choice but to enter a country with high cultural distance, managers can

reduce the negative impact of cultural distance by choosing the right entry modes.

Gatignon and Anderson (1988) stated that MNEs may require greater flexibility when the

cultural distance between the home and host country is high. Therefore, modes of entry with low

control and low risk, such as licensing, joint venture, merger and acquisition are preferred when it

comes to enter high cultural distance markets (Brouthers and Brouthers, 2003). These modes

allowed corporations to leverage the domestic firms’ experience, network of customers, suppliers

and distributors (Brouthers & Hennart 2007). It is crucial to keep in mind that cooperating with

local partners makes MNCs’ subsidiaries exposed to a threat of getting in conflict with local

stakeholders (Li & Shenkar, 1997). While licensing, joint venture, merger and acquisition can help

MNC to minimize the negative impact of cultural distance (Brouthers and Brouthers, 2003), when

expanding and growing to a certain level, MNCs must take full control of the business in order to

standardize the image and optimal the business performance (Hennart and Reddy, 1997). One way

to adapt better in this state is to put more effort into enhancing the efficiency of the operational

system. One outstanding example can be mentioned is Toyota, Monden (2011) described Taiichi

Ohno Toyota as a genius by inventing the Toyota Production System and Lean manufacturing. By

standardizing and automating the manufacturing with constant improvement the process, Toyota

factories are placed widely all over the world and yet they still grow rapidly without getting into

FTU Working Paper Series, Vol. 1 No. 2 (2/2022) | 241

conflict with other stakeholders in the host country (Spear & Bowen, 1999). In conclusion,

corporations should carefully take into account different aspects to choose the suitable strategy for

operating in a high cultural distance country.

On the other hand, several studies found a positive effect of cultural distance on the

performance of the subsidiary in general (Morosini et al., 1998). Managers can use this point of

view to leverage the potential benefit of learning from the host counterpart. This increases the

competencies, capabilities and even ideas for creative decision making situations which eventually

contribute to new competitive advantages for the parent corporation.

In summary, acknowledging the critical impact of cultural distance on performance of

employees, MNCs should carefully consider the difference between the MNCs home nationality

and Vietnam. In case Vietnam is so far the best option, having considered all aspects including

policy, economic distance, and other variables, high cultural distance MNCs with Vietnam should

have action plans to mitigate the impact with the above suggestion.

5.4. Limitations and suggestions for future research

Due to lack of time and resources, this paper is not without its limitations that should be taken

into consideration for further studies in the future.

First of all, even though the sample is representative of employee performance in MNCs in

Vietnam context, the modest sample size presents some problems regarding the power of statistical

analysis. According to Bryman & Bell (2015), to achieve random samples for conducting research

is a popular problem of research. Furthermore, although this paper also investigates employee

performance in different sectors, the data from the services sector contributes the most to the

collected data. Therefore, future studies may repeat the findings of this study on a larger sample

of the same population. And reducing the generalisability for service industries and diversifying

the sample with a variety of sectors should be made to ensure the analysis’s results.

Regarding employee performance scale, we want to concede that our study is based only on

self-report measures and there are differences according to the rater (Adler et al., 2016). Thus,

further research should analyse whether our findings are replicated with different raters, such as

supervisor or peers. We also suggest the study of content validity of the questionnaire applying

some coefficients such as Lawshe’s (1975) content validity ratio and Aitken’s (1980) coefficients

to give more evidence about its fit to the performance domain.

Finally, the current paper applies the cultural distance of Hofstede (2010) which is usually

employed in most research using cultural distance for analysis. Based on this dimension, the paper

can achieve a comprehensive evaluation of cultural distance on employee performance. However,

there is not enough measurement of cultural distance between Vietnam and some countries, such

as Vietnam and Cambodia. Future researchers may put more effort to utilise more specific

indicators to use a variety of cultural distance measurements to gain a more comprehensive

consideration.

FTU Working Paper Series, Vol. 1 No. 2 (2/2022) | 242