Quantification, Validation, And

8/12/2019 Quantification, Validation, And

1/113

THREE ESSAYS ON THE QUANTIFICATION, VALIDATION, AND

APPLICATION OF GRAYS ACCOUNTING VALUES

RAMON P. RODRIGUEZ, JR.

International Business Doctoral Program

APPROVED:

Gary P. Braun, Ph.D., Chair

Stephen B. Salter, Ph.D.

Patricia Eason, Ph.D.

Sidney J. Gray, Ph.D.

Patricia D. Witherspoon, Ph.D.Dean of the Graduate School


2/113

Copyright

by

Ramon P. Rodriguez, Jr.

2009


3/113

Dedication

To my wife and children. Thank you Rebecca, R.T., Robert, and Rex for your patience, love,

and support during my doctoral program years. You have made the difficult times easier and thegood times fantastic.

To Brian, Kate, Manuel, and Wayne. Your support and successes have motivated me to continuemy journey through tough times.

To Fernie R. Aceves. I could not have accomplished this without all of your help. You havebeen a professional mentor and I have learned more about the practice of accounting during mytime with you than during any other time in my career. You have been a wonderful father-in-law

providing love, guidance, and support to us. You have also been a treasured friend.


4/113


5/113

THREE ESSAYS ON THE QUANTIFICATION, VALIDATION, AND

APPLICATION OF GRAYS ACCOUNTING VALUES

by

RAMON P. RODRIGUEZ, JR., B.B.A, M.B.A.

DISSERTATION

Presented to the Faculty of the Graduate School of

The University of Texas at El Paso

in Partial Fulfillment

of the Requirements

for the Degree of

DOCTOR OF PHILOSOPHY

International Business Doctoral Program

THE UNIVERSITY OF TEXAS AT EL PASO

May 2009


6/113

UMI Number: 3358880

INFORMATION TO USERS

The quality of this reproduction is dependent upon the quality of the copy

submitted. Broken or indistinct print, colored or poor quality illustrations and

photographs, print bleed-through, substandard margins, and improper

alignment can adversely affect reproduction.

In the unlikely event that the author did not send a complete manuscript

and there are missing pages, these will be noted. Also, if unauthorized

copyright material had to be removed, a note will indicate the deletion.

______________________________________________________________

UMI Microform 3358880Copyright 2009 by ProQuest LLC

All rights reserved. This microform edition is protected againstunauthorized copying under Title 17, United States Code.

_______________________________________________________________

ProQuest LLC789 East Eisenhower Parkway

P.O. Box 1346Ann Arbor, MI 48106-1346


7/113

Acknowledgements

I would like to acknowledge the support and guidance of my doctoral committee

members Gary P. Braun, Ph.D., Stephen B. Salter, Ph.D., Patricia Eason, Ph.D., and Sidney J.

Gray, Ph.D. I also gratefully acknowledge Lance Eliot Brouthers, Ph.D., for giving me the

opportunity to show that I am worthy to be a member of this learned profession. I thank the PhD

Project, the KPMG Foundation, the American Institute of Certified Public Accountants

Fellowship for Minority Doctoral Students, the Deans Office, the Department of Accounting,

the Department of Marketing and Management, and other members of the College of Business

Administration at the University of Texas at El Paso as well as the School of Accountancy and

other members of the College of Business at Southern Illinois University Carbondale for

financial, intellectual, moral, and professional support during the completion of this dissertation.

Most importantly, I am forever indebted to my wife and children, my parents, my sister

and her family, my in-laws, and my friends for their constant support and love throughout my

careers, my studies and my life.

GO MINERS!

v


8/113

Abstract

This doctoral dissertation creates, validates and uses a quantification of country-level

accounting values based on Grays (1988) framework. This dissertation consists of three essays.

The first essay reports the theoretical foundation and method for the quantification of the

accounting values. The quantification results in reported scores for each of Grays (1988) four

accounting values Professionalism versus Statutory Control, Uniformity versus Flexibility,

Conservatism versus Optimism, and Secrecy versus Transparency. Accounting values scores are

reported for 58 countries. The second essay reports the method and results of validation testing

preformed on the accounting values scores. The primary validation method, Multi-Trait Multi-

Method Analysis, failed to provide empirical validation to the accounting values scores. A

secondary validation method, an ad-hoc research study replication, provided limited empirical

validation to the accounting values scores. The third essay uses the accounting values scores as a

primary variable in the explanation of differing levels of a countrys adoption of the International

Financial Reporting Standards. The essay reports empirical evidence of the effect of accounting

values on a countrys level of adoption of the International Financial Reporting Standards;

however, this effect seems to be disappearing as more countries adopt the International Financial

Reporting Standards.

vi


9/113

Table of Contents

Acknowledgements ..........................................................................................................................v

Abstract .......................................................................................................................................... vi

Table of Contents .......................................................................................................................... vii

List of Tables ................................................................................................................................. ix

List of Figures ..................................................................................................................................x

List of Illustrations ......................................................................................................................... xi

Overview ..........................................................................................................................................1Essay 1 - The Quantification of Grays Accounting Values............................................................4

1.1 Introduction ....................................................................................................................41.2 Theoretical Framework of Accounting and Culture ......................................................51.3 Hofstedes Cultural Dimensions ..................................................................................101.4 Grays Accounting Values ...........................................................................................12

1.4.1 Professionalism versus Statutory Control ..........................................................141.4.2 Uniformity versus Flexibility .............................................................................111.4.3 Conservatism versus Optimism .........................................................................151.4.4 Secrecy versus Transparency .............................................................................15

1.5

Quantification of Grays Accounting Values ..............................................................16

1.5.1 Gray (1988) and GLOBE (House et al. 2004) ...................................................21

1.6 Mechanics of the Quantification ..................................................................................241.6.1 Professionalism versus Statutory Control ..........................................................261.6.2 Uniformity versus Flexibility .............................................................................261.6.3 Conservatism versus Optimism .........................................................................271.6.4 Secrecy versus Transparency .............................................................................271.6.5 Standardization of Quantified Scores ................................................................281.6.6 Characteristics of Quantified Scores ..................................................................291.6.7 Empirical Test of Face Validity .........................................................................32

1.7 Discussion and Conclusion ..........................................................................................34Essay 2 - The Empirical Validation of the Accounting Values Scores .........................................37

2.1 Introduction ..................................................................................................................372.2 Multi-Trait Multi-Method Analysis .............................................................................38

2.2.1 Eddie (1990).......................................................................................................392.2.2 Salter and Niswander (1995) .............................................................................392.2.3 Gray and Vint (1995) .........................................................................................41

vii


10/113

2.2.4 Zarzeski (1996) ..................................................................................................412.2.5 La Porta et al. (1998) .........................................................................................432.2.6 Hope (2003) .......................................................................................................432.2.7 Analysis of Results ............................................................................................45

2.3 Replication of Jaggi and Low (2000)...........................................................................482.3.1 Variables and Regression Models ......................................................................492.3.2 Analysis of Results ............................................................................................54

2.4 Conclusion ...................................................................................................................61Essay 3 - Using Grays Accounting Values to Explain Deviations from IFRS .............................63

3.1 Introduction ..................................................................................................................633.2 Literature Review and Hypothesis Development ........................................................64

3.2.1 Conservatism versus Optimism .........................................................................653.2.2 Secrecy versus Transparency .............................................................................67

3.3

Research Method .........................................................................................................68

3.3.1 Variables and Regression Models ......................................................................683.3.2 Analysis of Results ............................................................................................743.3.3 Additional Test of H1 and H2 ............................................................................76

3.4 Conclusion ...................................................................................................................81Summary ........................................................................................................................................83

References ......................................................................................................................................87

Appendix 1 Composition of Eddies (1990) Accounting Values Indexes ..................................97

Appendix 2 Correlation of GLOBE and Hofstede Cultural Values Indices ...............................99

Vita.. .........................................................................................................................100

viii


11/113

List of Tables

Table 1.1: Matrix of Relationships of Accounting Values with Hofstede's Culture Values ........ 14

Table 1.2: Matrix of Relationships of Accounting Values with GLOBE Culture Values. ........... 22

Table 1.3: Accounting Practices Scores........................................................................................ 23

Table 1.4: Accounting Values Scores ........................................................................................... 25

Table 1.5: Descriptive Statistics, Test of Normality, and Pearson Correlations........................... 29

Table 1.6: Regression Results of Accounting Values Hypothesis Testing ................................... 33

Table 2.1: Multi-Trait Multi-Method Validity Results ................................................................. 46

Table 2.2: Descriptive Statistics on Regression Variables ........................................................... 52

Table 2.3: Jaggi and Low (2000) Replication Regression Results - Full Sample ........................ 55

Table 2.4: Jaggi and Low (2000) Replication Regression Results by Common Law and Code .. 58

Table 3.1: Dependent and Control Variables ................................................................................ 70

Table 3.2: Pearson Correlations for Variables in Tests on Absence Measures ............................ 72

Table 3.3: OLS Regression Results .............................................................................................. 75

Table 3.4: Pearson Correlations for Variables in Tests on Use of IFRS ...................................... 79

Table 3.5: Ordinal Probit Regression Results ............................................................................... 80

ix


12/113

List of Figures

Figure 1.1: Gray's (1988) Framework of Accounting and Culture. ................................................ 5

Figure 1.2: Perera's (1989) Framework of Accounting and Culture ............................................... 6

Figure 1.3: Doupnik and Tsakumis' (2004) Partial Refinement of Gray's (1988) Framework ofAccounting and Culture ................................................................................................ 7

Figure 1.4: Synthesis of Previous Frameworks into an institutional Framework of Accountingand Culture .................................................................................................................. 10

Figure 1.5: Histograms of Accounting Values Scores .................................................................. 31

x


13/113

xi

List of Exhibits

Exhibit 2.1: Composition of Salter & Niswanders (1995) Accounting Values Indexes ............. 40

Exhibit 2.2: Composition of Zarzeskis (1996) Disclosure Index ................................................ 42


14/113

Overview

This doctoral dissertation creates, validates and uses a quantification of country-level

accounting values based on Grays (1988) framework. Two decades after Gray (1988)

synthesized the extant literature to describe a framework of accounting values, a large scale

country-level quantification of accounting values still does not exist. A large scale country-level

quantification of accounting values would prove useful in a variety of studies. Many researchers

have used Grays accounting values framework as theoretical support for their studies in various

areas such as decision-making, firm valuation, materiality judgments, voluntary disclosure, and

foreign listing requirements (Adams et al. 1993; Archambault and Archambault 2003; Arnold et

al. 2001; Bao and Bao 2004; Buck and Shahrim 2005; Chow et al. 1995; Debreceny et al. 2002;

Doupnik and Richter 2003; Doupnik and Salter 1995; Hussein 1996; King and Langli 1998;

Lynn 1999; Meek et al. 1995; Radebaugh et al. 1995; Roberts and Salter 1999; Salter 1998;

Salter et al. 2001; Schultz and Lopez 2001); however, the accounting values proxies used by

these studies are problematic for one of two reasons either the study relies on the inclusion of

three or more cultural variables as a proxy for a single accounting value or it relies on a newly

created proxy that is not tested for validity nor reused in subsequent studies. The quantified

accounting values created in this dissertation would mitigate both of these issues by (1) replacing

multiple variables as a test of a single effect with one variable that tests a single effect, which

should aid in the interpretation of empirical results, and by (2) providing a uniform measure of

accounting values that can be used in all accounting values studies, which should aid in the

generalizability of empirical results. This dissertation consists of three essays in which a large

scale country-level quantification of Grays accounting values is created, tested, and applied.

1


15/113

The first essay reports the theoretical foundation and method for the quantification of the

accounting values. Although the quantification of the accounting values is based on Grays

(1988) framework, two other frameworks (Doupnik and Tsakumis 2004; Perera 1989) as well as

an institutional theory (DiMaggio and Powell 1991, 1983; Oliver 1991; Selznick 1948; Selznick

et al. 1969; Scott 1987, 2001) based synthesis of the three frameworks are presented and

discussed. The quantification consists of a summative combination of the GLOBE (House et al.

2004) cultural values index in the manner described in Grays (1988) accounting values

framework, which is then standardized to a range of 0 through 100 to aid with interpretation.

This quantification results in reported scores for each of Grays (1988) four accounting values

Professionalism versus Statutory Control, Uniformity versus Flexibility, Conservatism versus

Optimism, and Secrecy versus Transparency. Accounting values scores are reported for 58

countries. The results of ordinary least squares multiple regression analyses based on Grays

(1988) original four hypotheses provide empirical support for the face validity of the quantified

accounting values.

The second essay reports the method and results of validation testing preformed on the

accounting values scores. The primary validation method, Multi-Trait Multi-Method Analysis,

analyzes the correlations between the quantified accounting values scores and accounting values

proxies created in six previous studies. This validation method assesses construct validity1by

testing convergent2and discriminant

3validity. The Multi-Trait Multi-Method Analysis provided

only limited support for convergent validity and failed to provide empirical support for the

1Construct validity is the totality of evidence about whether a particular operationalization of a construct adequatelyrepresents what is intended by theoretical account of the construct being measured (Cronbach and Meehl 1955).2Convergent validity is based on the extent to which a measure correlates with different measures with which it istheoretically predicted to correlate (Kotz et al. 1985).3Discriminant validity is based on the extent to which an operationalization does not correlate withoperationalizations of different constructs (Kotz et al. 1985).

2


16/113

discriminant validity of the quantified accounting values scores. A secondary validation method,

an ad-hoc research study replication, provided limited empirical validation to the accounting

values scores.

The third essay uses the accounting values scores as a primary variable in the explanation

of differing levels of a countrys adoption of the International Financial Reporting Standards

(IFRS). While dozens of countries now require or permit the use of IFRS or are converging with

IFRS (International Accounting Standards Board 2009), the level of adoption varies between

countries (Dean and Clarke 2005). Ding et al. (2005) created an index of a countrys

nonconformity in its domestic GAAP with IFRS. Ding et al. (2005) reported the

nonconformities as an absence of a specific rule from domestic GAAP or as an instance of a

divergence of an existing domestic GAAP standard from the corresponding IFRS. I use Ding et

al.s (2005) measures of the absence of IFRS pronouncements from domestic GAAP as the

proxy for the differing levels of a countrys adoption of IFRS. I hypothesize and empirically

show that countries that value conservatism have higher levels of deviation from IFRS and that

countries that value secrecy have higher levels of deviation from IFRS. While this essay reports

empirical evidence of the effect of accounting values on a countrys level of adoption of IFRS,

an additional test of the hypotheses with newer data implies this effect may be disappearing as

more countries adopt the International Financial Reporting Standards.

3


17/113

Essay 1 - The Quantification of Grays Accounting Values

1.1 Introduction

Almost twenty years after Gray (1988) synthesized the extant literature to describe a

framework of accounting values, despite several partial studies (Chanchani and Willett 2004;

Eddie 1990; Salter and Niswander 1995), a large scale country-level quantification of accounting

values still does not exist. This is unlike the underlying framework for Gray (1988), i.e., the

cultural work of Hofstede (1980), which includes the creation of a country level quantification of

cultural dimensions and has been widely tested and replicated (Kirkman et al. 2006; Sondergaard

1994). In an attempt to address this gap in the accounting values literature, I develop scores for

each of the four accounting values described by Gray for approximately five dozen countries.

The process of quantifying Grays (1988) accounting values will assist future researchers

in examining the effect of accounting values on various accounting phenomena. For example,

many researchers have used Grays4 (1988; Radebaugh et al. 2006) accounting values

framework as theoretical support for their studies in various areas such as decision-making, firm

valuation, materiality judgments, voluntary disclosure, and foreign listing requirements (Adams

et al. 1993; Archambault and Archambault 2003; Arnold et al. 2001; Bao and Bao 2004; Buck

and Shahrim 2005; Chow et al. 1995; Debreceny et al. 2002; Doupnik and Richter 2003;

Doupnik and Salter 1995; Hussein 1996; King and Langli 1998; Lynn 1999; Meek et al. 1995;

Radebaugh et al. 1995; Roberts and Salter 1999; Salter 1998; Salter et al. 2001; Schultz and

Lopez 2001). A quantified set of accounting values scores would greatly assist in the hypothesis

testing of studies such as these.

4 In their textbook, Radebaugh, Gray, and Black (2006) extended the description of the framework presented byGray (1988).

4


18/113

1.2 Theoretical Framework of Accounting and Culture

Gray (1988) synthesized the work of others (Frank 1979; Mueller 1967, 1968; Nair and

Frank 1980; Nobes 1983, 1984) and presented a framework depicting the relationship between

accounting and culture (Figure 1.1). Perera (1989) proposes a modification of the Gray (1988)

framework (Figure 1.2). Additionally, in their review article, Doupnik and Tsakumis (2004)

presented a partial refinement of Grays (1988) framework (Figure 1.3). Doupnik and Tsakumis

(2004) specify Grays (1988) accounting systems construct as more than just country-level

financial reporting rules, but also the individual- and firm-level application of these rules by

accountants.

Figure 1.1: Gray's (1988) Framework of Accounting and Culture

5


19/113

Figure 1.2: Perera's (1989) Framework of Accounting and Culture

6


20/113

Figure 1.3: Doupnik and Tsakumis' (2004) Partial Refinement of Gray's (1988) Framework ofAccounting and Culture

The three frameworks (Doupnik and Tsakumis 2004; Perera 1989; Gray 1988) are similar

to each other in most respects. Each framework incorporates the influence of external forces on

the physical environment. Globalization and war, for example, affect societal-level physical or

ecological environments. These external forces cannot be controlled by the typical individual,

group, or society and can greatly influence the development of a society.

Each framework also describes the effect of societal institutions on accounting, but have

certain differences. Gray (1988) and Doupnik and Tsakumis (2004) show a complex relationship

of (1) an indirect relationship as societal culture affects institutions and institutions influence

accounting values and practices directly and (2) an additional direct relationship between

accounting values and practices. Perera (1989) also acknowledges societal culture but posits that

the effect of institutions on accounting values and practices will be simultaneous and separate.

7


21/113

The major difference between the three frameworks (Doupnik and Tsakumis 2004;

Perera 1989; Gray 1988) is the relationship between societal values and institutional

consequences. The Gray (1988) and Doupnik and Tsakumis (2004) frameworks illustrate a

hierarchical relationship where (1) societal values influence accounting values and institutional

consequences simultaneously and individually, and then (2) accounting values and institutional

consequences influence accounting practices simultaneously and individually. Alternatively,

Perera (1989) shows a direct effect of the physical environment on societal values and

institutional consequences as well as an interrelatedness between societal values and institutional

consequences. One possible solution to this difference is the application of institutional theory

(DiMaggio and Powell 1991, 1983; Oliver 1991; Selznick 1948; Selznick et al. 1969; Scott 1987,

2001).

Institutional theory describes the way in which organizations of all sizes and types

influence the behaviors of members of a society. Institutional attributes such as legal system,

political system, and religion affect not only the laws and norms of a society, but also the way its

members think. These three pillars (Scott 2001) of institutional theory (regulatory, normative,

and cognitive) encompass the institutional consequences construct of the Perera (1989) and Gray

(1988) frameworks as well as their societal values construct. The regulatory pillar includes

institutional consequences such as legal system, political system, capital markets, and corporate

ownership. The normative pillar includes institutional consequences such as professional

associations, education, and religion. The cognitive pillar includes societal values.

The different depictions of the antecedents and consequences of the institutional

consequences and societal values constructs in the Perera (1989), Gray (1988), and Doupnik and

Tsakumis (2004) frameworks are likely due to the complexities of the relationships and

8


22/113

processes described by institutional theory. Various institutional elements within a society can

jointly or individually affect one or more of the three pillars of institutional theory. An example

of the influence of societal values within the cognitive pillar of institutional theory is evident

when Sharp and Salter (1997) note that it is interesting to speculate whether agency theory

could have been developed to the extent it has in any culture other than an individualist one; the

very idea of rigorously modeling self-interest may not have occurred to researchers in a

collectivist culture.

The complex interactions of the two constructs (institutional consequences and societal

values) presented separately in the Perera (1989), Gray (1988), and Doupnik and Tsakumis

(2004) frameworks is better depicted as a single construct under the umbrella of institutional

theory. In other words, institutional theory reconciles the major difference in the frameworks.

Doupnik and Salter (1995) performed a similar combination in their general model of accounting

development. A combination of the Perera (1989), Gray (1988), and Doupnik and Tsakumis

(2004) frameworks is presented as an institutional framework of accounting and culture in Figure

1.4.

The institutional framework presented in Figure 1.4 depicts the influence of institutional

elements on accounting values and accounting practices, with an additional influence of

accounting values on accounting practices. These elements are diverse and broad. The list of

institutional elements in Figure 1.4 is not exhaustive; however, culture has been shown to proxy

for many of these. For example, Chui, Lloyd, and Kwok (2002) found a relationship between

culture and corporate capital structure. Additionally, Stulz and Williamson (2003) linked culture

to a countrys level of investor protection. I use culture as a proxy for the various institutional

influences on accounting values.

9


23/113

External Influences

Forces of natureTradeInvestmentConquest

Ecological Influences

GeographicEconomicDemographicGenetic/hygienicHistoricalTechnologicalUrbanization

Institutionalism

Regulatory, Normative, & Cognitive

Legal systemPolitical systemCapital marketsCorporate ownershipProfessional associationsEducationReligionSocietal Values

Accounting Values

ProfessionalismUniformityConservatismSecrecy

Accounting Practices

AuthorityApplication/EnforcementMeasurementDisclosure

Reinforcement

Figure 1.4: Synthesis of Previous Frameworks into an institutional Framework of Accountingand Culture

1.3 Hofstedes Cultural Dimensions

Gray (1988) based his accounting values framework largely on the cultural work of

Hofstede (1980). Hofstede (1980) described and quantified the structural elements of culture

that affect business. He presented these structural elements as four underlying dimensions of

societal culture. These four cultural dimensions are Large versus Small Power Distance, Strong

10


24/113

versus Weak Uncertainty Avoidance, Individualism versus Collectivism, and Masculinity versus

Femininity. Hofstede (1984: 83-84) described these four cultural dimensions as:

Large versus Small Power Distance Power Distance is the extent to which the

members of a society accept that power in institutions and organizations isdistributed unequally. This affects the behaviour of the less powerful as well as ofthe more powerful members of society. People in Large Power Distance societiesaccept a hierarchical order in which everybody has a place which needs no furtherjustification. People in Small Power Distance societies strive for powerequalization and demand justification for power inequalities. The fundamentalissue addressed by this dimension is how a society handles inequalities amongpeople when they occur. This has obvious consequence for the way people buildtheir institutions and organizations.

Strong versus Weak Uncertainty Avoidance Uncertainty Avoidance is the

degree to which the members of a society feel uncomfortable with uncertainty andambiguity. This feeling leads them to beliefs promising certainty and tomaintaining institutions protecting conformity. Strong Uncertainty Avoidancesocieties maintain rigid codes of belief and behaviour and are intolerant towardsdeviant persons and ideas. Weak Uncertainty Avoidance societies maintain amore relaxed atmosphere in which practice counts more than principles anddeviance is more easily tolerated. The fundamental issue addressed by thisdimension is how a society reacts on the fact that time only runs one way and thatthe future is unknown: whether it tries to control the future or to let it happen.Like Power Distance, Uncertainty Avoidance has consequences for the waypeople build their institutions and organizations.

Individualism versus Collectivism Individualism stands for a preference for aloosely knit social framework in society wherein individuals are supposed to takecare of themselves and their immediate families only. Its opposite, Collectivism,stands for a preference for a tightly knit social framework in which individualscan expect their relatives, clan, or other in-group to look after them in exchangefor unquestioning loyalty (the word "collectivism" is not used here to describe anyparticular political system). The fundamental issue addressed by this dimension isthe degree of interdependence a society maintains among individuals. It relates topeople's self-concept: I or we.

Masculinity versus Femininity Masculinity stands for a preference in societyfor achievement, heroism, assertiveness, and material success. Its opposite,Femininity, stands for a preference for relationships, modesty, caring for theweak, and the quality of life. This fundamental issue addressed by this dimensionis the way in which a society allocates social (as opposed to biological) roles tothe sexes.

11


25/113

Some societies strive for maximum social differentiation between the sexes. Thenorm is then that men are given the more outgoing, assertive roles and women thecaring, nurturing roles. As in all societies most institutions are populated by men.Such maximum-social-differentiation societies will permeate their institutionswith an assertive mentality. Such societies become "performance societies"

evident even from the values of their women. I have called these societies"masculine". (In the English language, "male" and "female" are used for thebiological distinctions between the sexes; "masculine" and "feminine" for thesocial distinction. A man can be feminine, but he cannot be female.)

Other societies strive for minimal social differentiation between the sexes. Thismeans that some women can take assertive roles if they want to but especially thatsome men can take relationship-oriented, modest, caring roles if they want to.Even in these societies, most institutions are populated by men (maybe slightlyless than in masculine societies). The minimum-social-differentiation societies incomparison with their opposite, the maximum-social-differentiation societies, will

permeate their institutions with a caring, quality-of-life orientated mentality. Suchsocieties become "welfare societies" in which caring for all members, even theweakest, is an important goal for men as well as women.

I have called such societies "feminine". "Masculine" and "feminine" are relativequalifications: they express the relative frequency of values which in principle arepresent in both types of societies. The fact that even modern societies can bedifferentiated on the basis of the way they allocate their social sex role is notsurprising in the light of anthropological research on non-literate, traditionalsocieties in which the social sex role allocation is always one of the essentialvariables. Like the Individualism-Collectivism dimension, the Masculinity-Femininity dimension relates to people's self-concept: who am l and what is mytask in life?

1.4 Grays Accounting Values

Gray (1988) described the four accounting values presented in Figure 1.4 in terms of their

relationship to Hofstedes (1980) cultural dimensions. He summarized these relationships

through the presentation of four untested hypotheses. Additionally, he described the strength of

the various relationships with the terms strong, less strong, and weak. I will use these

hypotheses and the three levels of relationship strengths to help create the accounting values

scores. Each accounting value and related hypothesis is discussed below.

12


26/113

1.4.1 Professionalism versus Statutory Control

Professionalism versus statutory control (referred to as professionalism) is defined as a

societys preference for the exercise of individual professional judgment and the maintenance of

professional self-regulation as opposed to compliance with prescriptive legal requirements and

statutory control (Gray 1988). This construct captures the importance of authority in a societys

accounting rules and regulations (Perera 1989). Accountants must exercise professional

judgment in the course of their duties and this accounting value captures the variation in the

extent of the use of professional judgment at the country level. Gray (1988) posited that

professionalism has the strongest link with individualism and uncertainty avoidance. He also

proposed a less strong link with power distance. Grays (1988) first hypothesis follows:

H1: The higher a country ranks in terms of individualism and the lower it ranksin terms of uncertainty avoidance and power distance, the more likely it is torank highly in terms of professionalism.

Subsequently, Radebaugh, Gray, and Black (2006) expanded on professionalism by

noting a weak relationship with masculinity and long term orientation. They state that

professionalism is negatively related with long term orientation and positively related to the

assertiveness aspect of masculinity. The relationships between professionalism and all of the

relevant cultural dimensions are presented in Table 1.1.

13


27/113

Table 1.1: Matrix of Relationships of Accounting Values with Hofstede's (1980) Culture Values

Accounting Value

Relationship

Strength

Uncertainty

Avoidance

Long-Term

Orientation

Power

Distance Individualism Masculinity

Strong Negative Positive

Less Strong Negative

Weak Negative Positive

Strong Positive NegativeLess Strong Positive

Weak

Strong Positive Positive

Less Strong Negative Negative

Weak

Strong Positive Positive Positive Negative


Weak

Blank squares indicate no relationship.

Professionalism vs.

Statutory Control

Uniformity vs.

Flexibility

Conservatism vs.

Optimism

Secrecy vs.

Transparency

1.4.2 Uniformity versus Flexibility

Uniformity versus flexibility (referred to as uniformity) is defined as a societys

preference for the enforcement of uniform accounting practices between companies and for the

consistent use of such practices over time as opposed to flexibility in accordance with the

perceived circumstances of individual companies (Gray 1988). This construct captures the

importance of enforcement in a societys accounting rules and regulations (Perera 1989).

Accounting rules and regulations are, to some extent, applied differently between industries and

over time. This accounting value captures these variations in the flexibility of accounting rules

and regulations across reporting entities as well as across time at the country level. Gray (1988)

posited that uniformity has the strongest link with uncertainty avoidance and individualism. He

also proposed a less strong link with power distance. The relationships between uniformity and

the cultural dimensions are presented in Table 1.1. Grays (1988) second hypothesis follows:

H2: The higher a country ranks in terms of uncertainty avoidance and powerdistance and the lower it ranks in terms of individualism, the more likely itis to rank highly in terms of uniformity.

14


28/113

1.4.3 Conservatism versus Optimism

Conservatism versus optimism (referred to as conservatism) is defined as a societys

preference for a cautious approach to measurement so as to cope with the uncertainty of future

events as opposed to a more optimistic, laissez-faire, risk-taking approach (Gray 1988). This

construct captures the importance of measurement in a societys accounting rules and regulations

(Perera 1989). Accounting rules and regulations govern the acceptable methods for measuring

and reporting the value of net assets and profits. The reported value of net assets and profits for

the same underlying transactions can vary by country based on different levels of cautiousness.

This accounting value captures variations in the level of cautiousness at the country level. Gray

(1988) posited that conservatism has the strongest link with uncertainty avoidance. He also

proposed a less strong link with individualism and masculinity. Grays (1988) third hypothesis

follows:

H3: The higher a country ranks in terms of uncertainty avoidance and the lowerit ranks in terms of individualism and masculinity, the more likely it is torank highly in terms of conservatism.

Subsequently, Radebaugh, Gray, and Black (2006) expanded on conservatism by noting a

strong relationship with long term orientation. They state that conservatism is positively related

with long term orientation. The relationships between conservatism and all of the relevant

cultural dimensions are presented in Table 1.1.

1.4.4 Secrecy versus Transparency

Secrecy versus transparency (referred to as secrecy) is defined as a societys preference

for confidentiality and the restriction of disclosure of information about the business only to

those who are closely involved with its management and financing as opposed to a more

transparent, open and publicly accountable approach (Gray 1988). This construct captures the

15


29/113

importance of disclosure in a societys accounting rules and regulations (Perera 1989). Financial

statements are meant to depict the state of the firm in an accurate, yet concise, way. This balance

between accuracy and conciseness results in variations in disclosure at the country level. Gray

(1988) posited that secrecy has the strongest link with uncertainty avoidance, power distance,

and individualism. He also proposed a less strong link with masculinity. Grays (1988) fourth

hypothesis follows:

H4: The higher a country ranks in terms of uncertainty avoidance and powerdistance and the lower it ranks in terms of individualism and masculinity,the more likely it is to rank highly in terms of secrecy.

Subsequently, Radebaugh, Gray, and Black (2006) expanded on secrecy by noting a

strong relationship with long term orientation. They state that secrecy is positively related with

long term orientation. The relationships between secrecy and all of the relevant cultural

dimensions are presented in Table 1.1.

1.5 Quantification of Grays Accounting Values

Although the quantification of accounting values on a scale as large (58 countries) as I

am proposing has not been previously completed, previous researchers have attempted this on a

smaller scale (one to 39 countries). Eddie (1990) and Salter and Niswander (1995) create

proxies for the accounting values from measures such as the wording within audit opinions and

the valuation of assets and liabilities. They then examine the association between these measures

of accounting values and Hofstedes (1980) cultural dimensions as described in Grays (1988)

four accounting value hypotheses. Eddie (1990) found statistically significant associations

between his measures of accounting values and Hofstedes (1980) cultural dimensions for all

thirteen of Grays (1988) hypothesized relationships using data from thirteen countries5. Salter

5Eddies (1990) results have been questioned (Chanchani and MacGregor 1999; Doupnik and Tsakumis 2004) dueto the lack of independent validation of his accounting values measures.

16


30/113

and Niswander found statistically significant associations for six of Grays (1988) thirteen

hypothesized relationships using data from twenty-seven countries. Others (Gray and Vint 1995;

Hope 2003; Jaggi and Low 2000; Wingate 1997; Zarzeski 1996) have used a countrys

disclosure index to focus specifically on the secrecy accounting value, finding full to partial

support for Grays (1988) hypothesized relationships using six to thirty-nine different countries.

These studies have quantified Grays (1988) accounting values in various ways, such as content

analyses of audit opinion letters (Salter and Niswander 1995), the presence or absence of a

professional exam (Salter and Niswander 1995), and content analyses of financial statements

(Sudarwan and Fogarty 1996; Zarzeski 1996). A drawback to a majority of these studies is the

subjectivity inherent in the method of quantification of the accounting values (Chanchani and

MacGregor 1999; Doupnik and Tsakumis 2004).

In addition to these smaller scale attempts at the quantification of accounting values, a

large scale attempt has been initiated. Chanchani and Willett (2004) present an accounting

values survey (AVS) meant to measure a countrys accounting values via financial statement

preparers and users. While they state that the measurement of accounting values is in progress

on a large scale, they have to date only presented results for India and New Zealand. One of

their reasons for choosing these countries is their use of the English language. This avoids a very

troublesome challenge of international business research survey translation. This problem

must be addressed before this large scale quantification of accounting values can continue.

I propose to create a large scale quantification of accounting values by treating each

accounting value as a multi-item measure. The items comprising each accounting value measure

are the cultural dimensions described by Grays (1988) four hypotheses. For example,

hypothesis 1 states that the higher a country ranks in terms of individualism and the lower it

17


31/113

ranks in terms of uncertainty avoidance and power distance, the more likely it is to rank highly in

terms of professionalism. Therefore, to create the measure of a countrys professionalism, I

propose to combine available measures of that countrys individualism, uncertainty avoidance,

and power distance dimensions of culture in the directions and strengths posited by Gray (1988).

This indirect method of quantification essentially creates a proxy for each accounting value using

culture as a basis.

Hofstedes (1980) cultural dimension scores would seem to be the likeliest candidate for

the basis of the quantified accounting values because Gray (1988) used Hofstedes definitions of

cultural dimensions to describe accounting values. However, a major limitation of Hofstedes

scores is the age of the data. Hofstede collected his data in the 1960s and 1970s. Although

Hofstede (2001) argues that his scores are valid because culture changes slowly over time,

Shenkar (2001) posits that cultures can substantially change over time and cultural stability is an

illusion, thereby eroding the utility of Hofstedes (1980) scores. This and other criticisms have

led some researchers (Baskerville 2003; McSweeney 2002b, 2002a; Schwartz 1992, 1994) to

argue that Hofstedes (1980) scores may no longer be completely valid. Schwartz (1994, 1999)

presents cultural dimension scores based on data collected in the late 1980s and early 1990s;

however, only 38 countries are represented. On the other hand, the GLOBE team present

cultural dimension scores for 616societies based on recently collected data (House et al. 2004).

I use the GLOBE data as an alternative to Hofstedes (1980) cultural dimension scores for the

basis of the accounting values scores due to the scope and age of the GLOBE data.

6The GLOBE team administered their surveys in 62 societies, but scores for the Czech Republic are not presenteddue to response bias problems. Additionally, the GLOBE team captures different, yet similar, dimensions for twosocieties in each of three nations (Germany, South Africa, and Switzerland). Although culture can vary within theborders of a nation, accounting rules and regulations do not; therefore, I have averaged the GLOBE scores for thesocieties residing within one nation. This brings the total number of countries in my study to 58.

18


32/113

Initially, Hofstede (1980) described culture through the use of four dimensions

uncertainty avoidance, power distance, individualism, and masculinity. Subsequently, he added

a fifth dimension long term orientation (Hofstede 2001). As the state of the literature has

progressed, the cultural dimensions have become more complex; for example, Schwartz (1994)

described seven culture level value types. The GLOBE project (2004) followed Schwartz

(1994) theory-driven approach to scale development (Hanges and Dickson 2004) and identified

nine cultural dimensions.

The GLOBE project is primarily based on the results of the survey of over 17,000 middle

managers in three industries: banking, food processing, and telecommunications, as well as

archival measures of country economic prosperity and the physical and psychological well-being

of the cultures studied (House et al. 2004). The GLOBE cultural dimensions are uncertainty

avoidance, power distance, future orientation, institutional collectivism, in-group collectivism,

gender egalitarianism, performance orientation, assertiveness, and humane orientation. The

cultural dimensions are described as (House et al. 2004: 30, 239, 282, 513, 569):

Uncertainty Avoidance is the extent to which a society, organization, or grouprelies on social norms, rules, and procedures to alleviate the unpredictability offuture events.

Power Distance is the extent to which a community accepts and endorsesauthority, power differences, and status privileges.

Future Orientationis the degree to which a collectivity encourages and rewardsfuture-oriented behaviors such as planning and delaying gratification.

Institutional Collectivism is the degree to which organizational and societalinstitutional practices encourage and reward collective distribution of resourcesand collective action.

In-Group Collectivismis the degree to which individuals express pride, loyalty,and cohesiveness in their organizations or families.

19


33/113

Gender Egalitarianism is the degree to which a collective minimizes genderinequality.

Performance Orientation reflects the extent to which a community encouragesand rewards innovation, high standards, excellence, and performance

improvement.

Assertiveness is the degree to which individuals are assertive, confrontational,and aggressive in their relationships with others.

Humane Orientation is the degree to which an organization or societyencourages and rewards individuals for being fair, altruistic, friendly, generous,caring, and kind to others.

The nine GLOBE cultural dimensions can be related back to Hofstedes (1980, 2001) five

dimensions. Three of them can be equated on a one to one basis. Hofstede s (1980, 2001)

uncertainty avoidance, power distance, and long term orientation are similar to uncertainty

avoidance, power distance, and future orientation; however, as shown in Appendix 2, there are

no statistically significant correlations between Hofstedes uncertainty avoidance, power

distance, and long term orientation and GLOBEs uncertainty avoidance, power distance, and

future orientation. Hofstedes (1980) individualism dimension has been split into institutional

collectivism and in-group collectivism. As presented in Appendix 2, individualism is not

statistically associated with in-group collectivism and negatively correlated with institutional

collectivism (R=-0.513; p=0.00). Hofstedes (1980) masculinity dimension has been

deconstructed into gender egalitarianism, performance orientation, assertiveness, and humane

orientation. As shown in Appendix 2, masculinity is not statistically associated with any of these

GLOBE variables.

The GLOBE study decomposes the collectivism dimension of culture into institutional

and in-group collectivism. The primary difference between the two is the level of analysis. The

level of analysis concept portrays differences between the perspectives of individuals, groups,

20


34/113

organizations, societies, countries, and global regions. For example, an examination of the

economy at a high level of analysis would yield data on trade imbalances and tariff impacts

while an examination at a low level of analysis would yield data on consumer behavior and local

demand. Institutional collectivism concerns a higher level of analysis, such as at the societal

level. On the other hand, in-group collectivism focuses on a lower level of analysis, such as at

the work group level of analysis. Accounting rules and regulations and, hence, accounting

values, are applied at the country level. Therefore, institutional collectivism is the more

appropriate dimension for the quantification of accounting values. However, an exception to this

is the professionalism accounting value. This accounting value is concerned with the level of

decision-making authority granted at the individual accountant level. Consequently, the in-group

collectivism dimension is more appropriate for the quantification of the professionalism

accounting value.

1.5.1 Gray (1988) and GLOBE (House et al. 2004)

Gray (1988) explained the influence of culture on accounting with an objective and

comprehensive framework. Additionally, he equated the accounting values framework to

Hofstedes (1980) cultural dimensions in four testable hypotheses. However, in describing the

relationships between the accounting values framework and Hofstedes cultural dimensions Gray

(1988) left the door open for the use of a different measure of culture because he described the

particular aspect of each dimension that related to the accounting values framework. I use

Grays (1988) additional information to equate accounting values to the GLOBE cultural scores,

which makes the following quantification of the accounting values possible. A matrix of these

relationships is presented in Table 1.2.

21


35/113

Table 1.2: Matrix of Relationships of Accounting Values with GLOBE Culture Values

Accounting Value

Relationship

Strength

Uncertainty

Avoidance

Future

Orientation

Power

Distance

Institutional

Collectivism

Strong Negative


Weak Negative

Strong Positive PositiveLess Strong Positive

Weak

Strong Positive Positive

Less Strong Positive

Weak

Strong Positive Positive Positive Positive

Less Strong

Weak

Accounting Value

Relationship

Strength

Performance

Orientation

In-Group

Collectivism

Humane

Orientation Assertiveness

Strong Negative

Less StrongWeak Positive

Strong

Less Strong

Weak

Strong


Weak

Strong

Less Strong Negative Negative

Weak

Gender Egalitarianism is not associated with any Accounting Value.

Blank squares indicate no relationship.

Uniformity vs.

Flexibility

Conservatism vs.

Optimism

Secrecy vs.

Transparency

Secrecy vs.

Transparency

Professionalism vs.

Statutory Control

Uniformity vs.

Flexibility

Conservatism vs.

Optimism

Professionalism vs.Statutory Control

Additionally, the institutional framework presented in Figure 1.4 depicts accounting

practices as well as accounting values. These practices can be equated to the accounting values

on a one to one basis. Authority, enforcement, measurement, and disclosure are associated with

professionalism, uniformity, conservatism, and secrecy, respectively (Gray 1988; Perera 1989;

Radebaugh et al. 2006). In addition to the societal values scores in the GLOBE study (House et

al. 2004), the authors present societal practices scores for all 61 societies. I have quantified the

accounting practices scores using the same method as the accounting values quantification

method described below. These accounting practices scores are presented in Table 1.3.

22


36/113


37/113

1.6 Mechanics of the Quantification

Each accounting value is a summative combination of the GLOBE cultural dimensions

listed in Table 1.2. More specifically, the scores for the cultural dimensions in Table 1.2 that

comprise each of the accounting values are added together to create a single score for each of the

four accounting values. In addition, I have applied a weighting system to the quantification

process. This weighting system is based on the accounting values literature. Gray (1988;

Radebaugh et al. 2006) used the terms strong, less strong, and weak to describe the relationships

between the cultural dimensions and the accounting values. In order to facilitate a weighted

combination of the multiple items comprising an accounting value, I have translated this into a

weight of three, two, and one for strong, less strong, and weak, respectively. In other words, a

relationship described as less strong carries twice the effect in my weighting method as a

relationship described as weak. Similarly, a relationship described as strong carries three times

the effect of a relationship described as weak. As an alternative, I used weights of four, two, and

one for strong, less strong, and weak (i.e., strong carried twice the weight of less strong and less

strong carried twice the weight of weak); however, the difference between the two weighting

systems was small or nonexistent7and I employ the three, two, one weighting method here. This

weighting method is straightforward and considers the varying strengths of the relationships

described by Gray (1988; Radebaugh et al. 2006). The weighting is explained in greater detail

for each accounting value below.

7I tested the two weighting methods using paired ttests of means analyses. The 4-2-1 and 3-2-1 weighting methodsproduced no statistically significant difference in means for the Uniformity versus Flexibility (t= 0.490, p=0.626)and Secrecy versus Transparency (t=-0.991, p=0.326) accounting values. The 4-2-1 method produced aProfessionalism versus Statutory Control accounting value with a mean of 45.32 and the 3-2-1 method produced amean of 45.81. This small difference (0.49) is statistically significant (t=2.137, p=0.037). The 4-2-1 methodproduced a Conservatism versus Optimism accounting value with a mean of 52.39 and the 3-2-1 method produced amean of 51.61. This small difference (0.78) is statistically significant (t=-3.902, p=0.000).The four accountingvalues produced with the 4-2-1 method were virtually perfectly correlated with the four accounting values producedwith the 3-2-1 method (Pearson R0.998).

24


38/113

Table 1.4: Accounting Values Scores

Country

Albania 31 82 67 78

Argentina 33 64 61 54

Australia 65 26 22 26

Austria 84 20 16 14

Bolivia 22 80 60 78Brazil 48 83 75 66

Canada 68 11 14 16

China 60 75 49 58

Colombia 27 69 70 56

Costa Rica 45 62 50 56

Denmark 80 15 0 9

Ecuador 21 81 79 72

Egypt 13 88 78 86

El Salvador 0 100 83 81

England 66 28 22 27

Finland 87 2 9 0

France 64 47 36 38

Georgia 33 47 56 50

Germany 87 23 14 21Greece 41 80 67 64

Guatemala 29 68 71 66

Hong Kong 62 55 51 60

Hungary 48 43 52 46

India 63 54 53 54

Indonesia 38 84 80 80

Iran 28 100 87 84

Ireland 66 31 27 31

Israel 54 33 34 33

Italy 51 56 61 55

Japan 82 26 38 39

Kazakhstan 56 36 32 38

Kuwait 39 79 65 78

Malaysia 36 70 67 68

Mexico 17 80 74 77Morocco 15 89 83 84

Namibia 15 61 65 63

Netherlands 100 2 11 17

New Zealand 29 38 31 59

Nigeria 16 90 86 72

Philippines 28 70 68 66

Poland 39 49 34 44

Portugal 50 58 46 46

Qatar 31 81 71 81

Russia 25 40 54 44

Singapore 62 43 42 45

Slovenia 47 51 44 43

South Africa 43 53 51 58

South Korea 54 27 54 43Spain 49 62 67 55

Sweden 69 0 1 4

Switzerland 97 12 4 9

Taiwan 26 93 70 78

Thailand 6 96 100 100

Turkey 36 64 77 68

USA 65 21 19 24

Venezuela 17 83 77 69

Zambia 50 49 57 49

Zimbabwe 42 60 61 62

Professionalism vs.

Statutory Control

Uniformity vs.

Flexibility

Conservatism vs.

Optimism

Secrecy vs.

Transparency

25


39/113

1.6.1 Professionalism versus Statutory Control

As shown in Table 1.2, the professionalism accounting value is strongly and negatively

associated with uncertainty avoidance, weakly and negatively associated with future orientation,

less strongly and negatively associated with power distance, strongly and negatively associated

with in-group collectivism, and weakly and positively associated with assertiveness. To

illustrate the quantification, the accounting values scores for Australia will be calculated. Using

the weights described earlier and the GLOBE study (House et al. 2004) cultural dimension

scores, the professionalism score for Australia is the weighted sum of strong and negative

uncertainty avoidance (3 X -3.98), weak and negative future orientation (1 X -5.15), less strong

and negative power distance (2 X -2.78), strong and negative in-group collectivism (3 X -5.75),

and weak assertiveness (1 X 3.81). This score is standardized and reported in Table 1.4 with the

professionalism scores of the other 57 countries. Descriptive and other statistics for this score are

presented later in Panel B of Table 1.5.

1.6.2 Uniformity versus Flexibility

As shown in Table 1.2, the uniformity accounting value is strongly and positively

associated with uncertainty avoidance, less strongly and positively associated with power

distance, and strongly and positively associated with institutional collectivism. Using the

weights described earlier and the GLOBE study (House et al. 2004) cultural dimension scores,

the uniformity score for Australia is the weighted sum of strong uncertainty avoidance (3 X

3.98), less strong power distance (2 X 2.78), and strong institutional collectivism (3 X 4.40).

This score is standardized and reported in Table 1.4 with the uniformity scores of the other 57

countries. Descriptive and other statistics for this score are presented later in Panel B of Table

1.5.

26


40/113

1.6.3 Conservatism versus Optimism

As shown in Table 1.2, the conservatism accounting value is strongly and positively

associated with uncertainty avoidance, strongly and positively associated with future orientation,

less strongly and positively associated with institutional collectivism, and less strongly and

negatively associated with performance orientation. Using the weights described earlier and the

GLOBE study (House et al. 2004) cultural dimension scores, the conservatism score for

Australia is the weighted average of strong uncertainty avoidance (3 X 3.98), strong future

orientation (3 X 5.15), less strong institutional collectivism (2 X 4.40), less strong and negative

performance orientation (2 X -5.89). This score is standardized and reported in Table 1.4 with

the conservatism scores of the other 57 countries. Descriptive and other statistics for this score

are presented later in Panel B of Table 1.5.

1.6.4 Secrecy versus Transparency

As shown in Table 1.2, the secrecy accounting value is strongly and positively associated

with uncertainty avoidance, strongly and positively associated with future orientation, strongly

and positively associated with power distance, strongly and positively associated with

institutional collectivism, less strongly and negatively associated with performance orientation,

and less strongly and negatively associated with humane orientation. Using the weights

described earlier and the GLOBE study (House et al. 2004) cultural dimension scores, the

secrecy score for Australia is the weighted sum of strong uncertainty avoidance (3 X 3.98),

strong future orientation (3 X 5.15), strong power distance (3 X 2.78), strong institutional

collectivism (3 X 4.40), less strong and negative performance orientation (2 X -5.89), and less

strong and negative humane orientation (2 X 5.58). This score is standardized and reported in

27


41/113

Table 1.4 with the secrecy scores of the other 57 countries. Descriptive and other statistics for

this score are presented later in Panel B of Table 1.5.

1.6.5 Standardization of Quantified Scores

The GLOBE cultural dimension scores were gathered with a survey using a seven-point

scale (House et al. 2004). This characteristic was transferred to the unstandardized accounting

values scores8. For increased clarity of understanding and the intuitiveness of the accounting

values scores, the scores presented are standardized to create a range between zero and 100.

Standardization converts variables to a common scale and is generally performed to create data

with a mean of zero and standard deviation of one (Hair et al. 2006). For the accounting values

scores, I standardized to a given range rather than a given mean and standard deviation. This

creates scores that look similar to one another, but allows for different, albeit similar, means and

standard deviations. This is important because readers will likely be more comfortable reading a

secrecy score for Australia of 26 rather than 4.465. Furthermore, comparing the secrecy scores

of Australia and China may be easier for some if the scores are 26 and 58 rather than 4.465 and

4.780. The unstandardized and standardized scores are perfectly correlated, have the same

distribution, and are statistically identical in all other respects.

8The unstandardized accounting values scores are not presented.

28


42/113

Table 1.5: Descriptive Statistics, Test of Normality, and Pearson Correlations

Panel A: Accounting Practices Scores

Accounting Value Mean

Std.

Dev. Min. Max.

Kolmogorov-

Smirnov Z

(Asymp. Sig.)

Correl.

1.

(Sig.)

Correl.

2.

(Sig.)

Correl.

3.

(Sig.)

1. Professionalism vs. 57.89 21.86 0.00 100.00 0.641 Statutory Control (0.805)

(Authority)

2. Uniformity vs. 54.94 21.18 0.00 100.00 0.416 0.404

Flexibility (0.995) (0.002)

(Enforcement)

3. Conservatism vs. 61.45 22.94 0.00 100.00 0.753 0.162 0.883

Optimism (0.621) (0.224) (0.000)

(Measurement)

4. Secrecy vs. 64.55 21.16 0.00 100.00 1.091 0.277 0.857 0.887

Transparency (0.184) (0.036) (0.000) (0.000)

(Disclosure)

n = 58

Panel B: Accounting Values Scores

Accounting Value Mean

Std.

Dev. Min. Max.

Kolmogorov-

Smirnov Z

(Asymp. Sig.)

Correl.

1.

(Sig.)

Correl.

2.

(Sig.)

Correl.

3.

(Sig.)

1. Professionalism vs. 45.81 23.33 0.00 100.00 0.527

Statutory Control (0.944)

2. Uniformity vs. 54.96 27.15 0.00 100.00 0.770 -0.842

Flexibility (0.593) (0.000)

3. Conservatism vs. 51.61 24.85 0.00 100.00 0.785 -0.859 0.935 Optimism (0.568) (0.000) (0.000)

4. Secrecy vs. 52.47 23.56 0.00 100.00 0.663 -0.865 0.955 0.946

Transparency (0.772) (0.000) (0.000) (0.000)

n = 58

Panel C: Correlations between Accounting Practices Scores and Accounting Values Scores

Professionalism vs.

Statutory Control

(Sig.)

0.203

(0.126)

n = 58

(0.001) (0.000) (0.000)

Uniformity vs.

Flexibility

(Sig.)

Conservatism vs.

Optimism

(Sig.)

Secrecy vs.

Transparency

(Sig.)

0.5190.6230.425

1.6.6 Characteristics of Quantified Scores

Table 1.5 presents some descriptive information about the quantified scores. The

accounting practices scores in Table 1.3 do not have a theoretical foundation; instead they are the

29


43/113

result of accounting values theory applied to cultural practices data. They are presented for

informational purposes and the associated statistics presented in Panel A of Table 1.5 and later in

Figure 1.5 are not discussed. The accounting values scores have means of 46 to 55 and standard

deviations of 23 to 27, as presented in Panel B of Table 1.5. Due to the standardization step of

the quantification process, all of the scores range from zero to 100. A normal distribution is an

important characteristic in data such as these because of the assumptions of ordinary least

squares regression. Table 1.5 Panel B shows the Kolmogorov-Smirnov Z scores for all four

accounting values scores. All four of the accounting values scores have Kolmogorov-Smirnov Z

scores that are not statistically significant. This indicates normal distributions for all four scores

(Massey 1951). Hair et al. (2006) suggest the use of graphical plots in conjunction with

statistical tests. Accordingly, Figure 1.5 presents histograms of all four accounting values scores

with normal distribution curves. These also appear to be acceptably, albeit not perfectly,

normally distributed.

30


44/113

Figure 1.5: Histograms of Accounting Values Scores

31


45/113

Table 1.5 Panel B also presents the correlations between the four accounting values

scores. The absolute values of the correlations are in the range of 0.842 to 0.955. These high

correlations are caused by using many of the same eight cultural dimensions to create each of the

accounting values scores in a summative manner. For example, uncertainty avoidance is

strongly associated with all four accounting values. This one cultural dimension makes up at

least 19% (3/16 of secrecy) and as much as 38% (3/8 of uniformity) of each of the accounting

values scores. These high correlations could cause multicollinearity in regression models where

two or more of the quantified accounting values are in the same model. However, the scores still

have utility when used separately. Many studies focus on a single accounting value at a time

(e.g., Debreceny et al. 2002; Gray and Vint 1995; Hope 2003; Jaggi and Low 2000; King and

Langli 1998; Meek et al. 1995) and the accounting values scores would work well in these

situations.

1.6.7 Empirical Test of Face Validity

Grays (1988) accounting value hypothesis were stated in terms of Hofstedes (1980)

cultural measures and the GLOBE (House et al. 2004) study provides alternate measures of

culture; therefore, an empirical test of the relationship between these accounting values scores

and Hofstedes cultural dimension scores using Grays (1988) four hypotheses may appear to

simply be a test of the similarity of the GLOBE and Hofstede measures of cultural dimensions.

However, I believe it is a necessary initial test of the accounting values scores in order to provide

face validity. I test these hypotheses using regression analysis on four models where each of the

four accounting value scores is explained by the cultural dimensions in each of Grays (1988)

four hypotheses. For example, hypothesis 1 is tested by examining the relationship between the

professionalism score and Hofstede s (1980) individualism, uncertainty avoidance, and power

32


46/113

distance scores. There are 49 countries9with both GLOBE and Hofstede values reported. These

49 countries comprise the sample for the regression analysis.

Table 1.6: Regression Results of Accounting Values Hypothesis Testing

Pred.

Professionalism

vs.

Statutory Control

Pred.

Uniformity

vs.

Flexibility

Pred.

Conservatism

vs.

Optimism

Pred.

Secrecy

vs.

Transparency

Individualism + 0.317 - -0.449 - -0.717 - -0.424(0.02) (0.00) (0.00) (0.00)

- -0.208 + 0.124 + 0.166 + 0.086(0.03) (0.11) (0.05) (0.20)

- -0.351 + 0.345 + 0.377(0.01) (0.01) (0.00)

Masculinity - 0.143 - 0.117(0.07) (0.12)

R-Square 0.472 0.586 0.600 0.581

Adj. R-Sqr. 0.437 0.558 0.573 0.543

F-Statistic 13.400 21.240 22.470 15.245

F-Stat. Sig. 0.000 0.000 0.000 0.000

n = 49

Standardized coefficients are presented.

Significance levels are presented in parentheses below coefficients.

All hypothesis tests are one-tailed.

F tests of the models are two-tailed.

Hypothesis 3 Hypothesis 4

Uncertainty

Avoidance

Power

Distance

Hypothesis 1 Hypothesis 2

Table 1.6 presents the tests of the thirteen relationships posited in Grays (1988) four

hypotheses. These tests statistically support nine of the thirteen relationships. The tests of the

relationships comprising the professionalism and uniformity accounting values are all

statistically significant at the p=0.05 level or better, with the exception of the relationship

between uniformity and uncertainty avoidance, which was in the predicted direction but has a p-

value of 0.11. This provides an initial level of validity to these two accounting values scores.

9 The 49 common countries are Argentina, Australia, Austria, Brazil, Canada, China, Colombia, Costa Rica,Denmark, Ecuador, Egypt, El Salvador, England, Finland, France, Germany, Greece, Guatemala, Hong Kong,Hungary, India, Indonesia, Iran, Ireland, Israel, Italy, Japan, Kuwait, Malaysia, Mexico, Netherlands, New Zealand,Nigeria, Philippines, Poland, Portugal, Russia, Singapore, South Africa, South Korea, Spain, Sweden, Switzerland,Taiwan, Thailand, Turkey, USA, Venezuela, and Zambia.

33


47/113

The relationship between secrecy and uncertainty avoidance is in the predicted direction;

however, it failed to receive statistical support (p=0.20). The masculinity dimension is not

statistically related to either the conservatism or secrecy accounting values. This is likely due to

the use of the decomposed aspects of the masculinity dimension, specifically assertiveness,

performance orientation, and humane orientation, instead of Hofstedes (1980) masculinity

score. However, the other relationships comprising the conservatism and secrecy accounting

values are all in the predicted direction and, with one exception, statistically significant at the

p=0.05 level or better. Overall, these tests provide, at the very least, face validity to the

accounting values scores.

1.7 Discussion and Conclusion

There are several limitations in my quantification of the accounting values. Among these

is the assumption of the validity of Grays (1988) accounting values. Doupnik and Tsakumis

(2004) review numerous empirical tests of Grays (1988) framework and conclude that the

framework has not been subjected to adequate empirical inquiry to positively support it

empirically. These empirical tests of Grays (1988) framework include Salter and Niswander

(1995), who found empirical support for only six of the thirteen relationships posited in Grays

(1988) four hypotheses. Sudarwan and Fogarty (1996) found empirical support for only four of

the relationships. While there are other studies that support Grays (1988) framework, Eddie

(1990) found empirical support for all thirteen hypothesized relationships, there is still enough

ambiguity to withhold a judgment of final empirical support for the Gray (1988) framework.

A second limitation is the appearance of circular logic in the quantification and validation

of the accounting values. Grays (1988) hypotheses are used to quantify the accounting values

and the same relationships are used to provide initial validation to the quantified accounting

34


48/113

values. A solution to this apparent circular logic would be quantifying the accounting values

using primary data gathered from hundreds of subjects from each of dozens of countries. This

solution would be an ambitious and worthwhile undertaking, but in the interim my quantification

using secondary data provides a necessary second-best solution.

Another limitation is the coverage of countries. An advantage to using secondary data to

quantify the accounting values is the large number of countries included in the quantification.

However, this number is limited by the secondary data used in the quantification. In this case,

the GLOBE data covers only 58 of the worlds 194 countries. Furthermore, the initial validity

testing of the quantified accounting values is done on only 49 countries due to the limited

overlap of the GLOBE and Hofstede data. This leaves three-quarters of the worlds countries out

of the process. The future availability of a more inclusive cultural dataset would be the best

solution, but a valid combination the GLOBE (House et al. 2004), Hofstede (1980), and

Schwartz (1994, 1999) datasets could substantially increase the coverage of countries. This is

left for future endeavors.

In this essay I have presented an institutional framework of the relationship between

accounting and culture. This framework integrates institutional theory with prior accounting

values frameworks (Doupnik and Tsakumis 2004; Gray 1988; Perera 1989). I have also

quantified Grays (1988) accounting values for 58 countries. While attempts have been

previously made to quantify Grays (1988) accounting values with survey instruments (e.g.,

Chanchani and Willett 2004), this work has not yet been completed. As an alternative to

collecting primary data on accounting values, I have used theory formulated by Gray (1988)

from the work of others (Frank 1979; Mueller 1967, 1968; Nair and Frank 1980; Nobes 1983,

1984) and cultural dimensions quantified by the GLOBE project (House et al. 2004) to enrich the

35


49/113

accounting values research stream by presenting quantified accounting values scores for the

purpose of expanding and testing this theoretical realm by future researchers.

The next step in this process is to validate these quantified accounting values scores. I do

this in Essay 2 with standard validation techniques and through the use of multiple replications

of previous studies. Essay 3 applies the quantified accounting values scores to an examination of

deviations of a countrys domestic GAAP from International Financial Reporting Standards.

36


50/113

Essay 2 - The Empirical Validation of the Accounting Values Scores

2.1 Introduction

With the creation of the accounting values scores in Essay 1 comes the need to test the

validity of these scores in a variety of ways. The accounting values scores will provide only

limited help to international accounting researchers unless they are shown to be a reliable proxy

for the accounting values construct. A traditional validity test as well as a replication of a

previous study using Grays (1988) framework will be undertaken in an attempt to validate the

accounting values scores.

The accounting values scores created in Essay 1 will be tested using two different

methods. First, previous research that has attempted to test Grays (1988) framework (e.g.,

Eddie 1990; Gray and Vint 1995; Hope 2003; Salter and Niswander 1995) will be used as

various methods of construct measurement in a Multi-Trait Multi-Method validation analysis.

The accounting values scores created in Essay 1 are tested against various proxies for accounting

values created in these previous studies. The expectation is that quantified accounting values

created in Essay 1 will be properly related to the various other proxies for accounting values.

For example, Salter and Niswander (1995) created multiple measures of each of the four

accounting values and tested these measures against Hofstedes (1980) cultural dimension scores

using Grays (1988) hypothesized relationships. The Multi-Trait Multi-Method analysis will use

the Salter and Niswander (1995) measures, along with others, to provide validity to the

accounting values scores created in Essay 1.

Second, other researchers have used Grays (1988) framework as theoretical support for

their studies (e.g., Arnold et al. 2001; Doupnik and Riccio 2006; Jaggi and Low 2000; Meek et

al. 1995; Schultz and Lopez 2001). The second method for validating the accounting values

37


51/113

scores involves replicating one of these studied by replacing their proxy for culture with the

appropriate quantified accounting values score. The expectation is that quantified accounting

values created in Essay 1 will improve the explanatory value of each model and, therefore,

improve the hypothesis testing of the studies.

The results of these two validation methods are mixed. The Multi-Trait Multi-Method

analysis shows limited evidence of convergent validity and no evidence of discriminant validity.

Conversely, the replication method shows evidence of the usefulness of the quantified

accounting values scores. Taken together, I reach the conclusion that, although the accounting

values scores are not as statistically sound, they do provide help in addressing research questions

dealing with culture and international accounting practices and systems.

2.2 Multi-Trait Multi-Method Analysis

Multi-Trait Multi-Method analysis was introduced by Campbell and Fiske (1959) and

tests construct validity using correlation analysis between measurements obtained when each of

a number of traits is measured by each of a number of methods (Kotz et al. 1985). Cronbach and

Meehl (1955) define construct validity as the totality of evidence about whether a particular

operationalization of a construct adequately represents what is intended by theoretical account of

the construct being measured. Construct validity is primarily comprised of convergent and

discriminant validity. Convergent validity is based on the extent to which a measure correlates

with different measures with which it is theoretically predicted to correlate and discriminant

validity is based on the extent to which an operationalization does not correlate with

operationalizations of different constructs (Kotz et al. 1985). As a test of the construct validity

of the accounting values created in Essay 1, I use Multi-Trait Multi-Method analysis to compare

similarities and differences with accounting value proxies created in the following six studies.

38


52/113

2.2.1 Eddie (1990)

Eddie (1990) constructed indexes of accounting values and used them to test the link

between a countrys societal values and characteristics of its accounting system. He selected the

factors comprising the indexes on the basis of theoretical reasoning (Eddie 1990). These 40

factors, ten factors per accounting value, are listed in the Appendix of this dissertation. Similar

to Nobes (1984), information for scoring the factors for each country was discerned from works

of reference, study of financial reports and general business literature (Eddie 1990). Each factor

was then scored judgmentally on a six point scale (Eddie 1990). Eddie (1990) used correlation

analysis to empirically support all of Grays (1988) hypothesized relationships.

Eddies (1990) accounting values scores are presented in his paper. The data covers

thirteen countries, twelve of which overlap with the accounting values scores created in Essay 1.

This data is method 2 in the Multi-Trait Multi-Method analysis.

2.2.2 Salter and Niswander (1995)

Salter and Niswander (1995) also created scores for Grays (1988) accounting values in

order to test the

Quantification, Validation, And

Documents

Transcript of Quantification, Validation, And