artikel fgd

8/17/2019 artikel FGD

1/27

JOURNAL OF INTERNATIONAL ACCOUNTING RESEARCH Vol. 7, No. 12008pp. 25–50

Voluntary Accounting Disclosures byU.S.-Listed Asian Companies

Gaurav Kumar, W. Mark Wilder, and Morris H. Stocks

ABSTRACT: The current study examines the voluntary disclosures (provided in the

U.S.) by U.S.-listed Asian companies. Our findings indicate that significantly fewer

(greater) voluntary disclosures are provided by U.S.-listed Asian companies from coun-

tries which have a strict (less strict) mandatory disclosure regime in their home country.

This finding is contrary to the model developed in Einhorn 2005. In addition, this studydocuments that over 80 percent of U.S.-listed Asian companies are voluntarily using

‘‘international’’ standards for their financial statements and thus contributing towards

international accounting convergence. However, their choice to use ‘‘international’’

standards is not affected by their proportion of foreign sales or size. An important

contribution of this study is the development of a measure for strictness of mandatory

disclosure regimes of Asian countries by using expert rankings.

I. INTRODUCTION AND MOTIVATION

The purpose of this paper is to collect empirical evidence on the voluntary disclosuresprovided (in the U.S.) by U.S.-listed Asian companies (hereafter, referred to asUSLAC). Various studies have examined the voluntary disclosures of companies in

a domestic context (e.g., Botosan 1997) as well as internationally (examples includeZarzeski [1996] and Cahan et al. [2005]). However, an area of the literature that has notbeen well explored is the voluntary disclosures of USLAC. This paper examines three issuesin this context. The first issue analyzes the effect of home country mandatory disclosureenvironment on the voluntary disclosures provided by USLAC in the U.S. The second issueexamines the extent to which the voluntary disclosures of sample companies are convergentand the effect of domestic and ‘‘global’’ culture on those disclosures. The third issue in-vestigates the voluntary use of ‘‘international’’ standards instead of national standards bythe sample companies in preparing their consolidated financial statements.

Einhorn (2005) indicates that most of the literature on voluntary disclosures ignoresthe existence of the mandatory disclosure environment. She examines the interaction be-tween a firm’s mandatory and voluntary disclosures, and develops a model which indicatesthat a firm’s voluntary disclosure strategy is crucially affected by the presence of its ownmandatory disclosures. Einhorn (2005) specifically shows that there should be a positiverelationship between the mandatory and voluntary disclosures made by a firm, meaningmore stringent mandatory disclosure requirements will lead to enhanced voluntary disclo-sures. She defines a stringent mandatory disclosure environment as one where there is lessdiscretion for companies in designing their periodic reports and reasons that more discretionin mandatory reporting enlarges the manager’s set of disclosure alternatives. This additional

Gaurav Kumar is an Assistant Professor at the University of Arkansas at Little Rock, and W Mark Wilder and Morris H Stocks are both Professors at The University of Mississippi


2/27

26 Kumar, Wilder, and Stocks

discretion compared to other alternatives could result in minimizing the need for voluntarydisclosures.

Mandatory disclosures refer to the content, in the periodic financial reports, that com-panies are required to submit. The types of reports that are required typically are the same

in various countries. For example, annual reports and interim half-yearly statements areordinarily required in every country. Some countries require quarterly financial statementsfrom the listed companies as well. However, some countries have very stringent rules forthe content of these reports whereas other countries allow more flexibility in reporting.1

The information beyond the required content in the financial statements is voluntary innature. Addressing the potential for variety in voluntary disclosures, Lang and Lundholm(1993) state:

While the SEC mandatory disclosure requirements provide a basic framework and minimumstandard for many financial disclosures, considerable latitude remains in determining what infor-mation is actually provided. Some firms’ annual and quarterly reports go well beyond the requireddisclosures, while others are extremely stark.

USLAC provide a good setting for testing the validity of the Einhorn (2005) modelbecause these companies belong to different mandatory disclosure environments. Moreover,Asian companies are used in this study due to their growing importance in the worldeconomy. The sample companies, listed in the U.S., from these Asian countries are someof the best performers in their respective industries (e.g., Canon, Honda Motors, SonyCorporation, and Toyota Motors). In addition, some Asian countries (e.g., China and India)have the fastest growing economies in the world. For example, a recent article in USAToday’s Money section indicated that China’s economy is growing roughly three times asfast as the USA’s. The article further states that U.S.-listed Chinese companies have seentheir shares increase 72 percent on average during the first three quarters of 2007 (Krantz2007).

This paper examines the Einhorn (2005) model in the context of voluntary accountingdisclosures (in the U.S.) by USLAC in the years 2003 and 2004. The mandatory disclosurerequirements in some Asian countries are difficult to differentiate (examples include Chinaand Hong Kong) by just examining the requirements. As a result, we requested experts inthe area of international accounting research (specifically in the Asian-Pacific area) torank the sample countries based on the strictness of their mandatory disclosure regime. Forpurposes of testing the Einhorn (2005) model, voluntary disclosures of sample companiesare measured using the approach in Botosan (1997).2

Another related research question examined in this paper is the extent of harmonization

or convergence of voluntary disclosure practices by USLAC. Warner (2003) shows that allAsian countries have a distinct cultural framework and in many cases a form of managementwith local characteristics. More importantly, Warner (2003) reports that in Asia, country-specific cultures have resulted in an observable set of highly identifiable institutions leadingto different management styles in all countries. Therefore, it can be argued that sincedomestic culture is an important determinant for management styles in Asian countries, itwill also affect the accounting disclosures by USLAC.

1 In the current research, ‘‘more stringent’’ refers to more quantity and more intense disclosure of informationitems, as measured under home stock exchange disclosure requirements. Quantity refers to the number of information items to be disclosed, while intensity refers to the degree of detail in which information items arerequired to be disclosed (Adhikari and Tondkar 1992).

2 Botosan (1997) examined the association between cost of equity capital and voluntary disclosure level. Shedeveloped her own index to measure the voluntary disclosures of her sample firms. This index has been widelyused in the literature to measure voluntary disclosures


3/27

Voluntary Accounting Disclosures by U.S.-Listed Asian Companies 27

On the other hand, Zarzeski (1996) argues that foreign-listed companies tend to borrowa ‘‘global’’ culture which results in their disclosures becoming similar across firms. Shecompares the disclosure practices of 256 companies (domestic-only and foreign-listed) fromseven countries and reports that culture is an important variable that affects the disclosures

of domestic-only companies. However, she also reports that when companies become listedin foreign countries, the effect of home culture on their disclosures decreases because thosecompanies tend to borrow a ‘‘global’’ culture. Based on these arguments, we compare thevoluntary disclosures by all USLAC in the U.S., measured by Botosan (1997), to examinethe effect of home culture and ‘‘global’’ culture. Results showing a borrowing of ‘‘global’’culture would provide evidence on convergence and vice-versa.

Lastly, we investigate the voluntary use of ‘‘international’’ (U.S. Generally AcceptedAccounting Principles [GAAP] and International Financial Reporting Standards [IFRS])standards by USLAC in their Form 20-F annual report. Tarca (2004) examines the disclo-sure practices of foreign and domestic-only companies in the U.K., France, Germany, Japan,and Australia and reports that companies using ‘‘international’’ standards are likely to be

bigger in size and have more foreign revenue. Therefore, this paper examines the signifi-cance of firm size and foreign revenue in the voluntary use of ‘‘international’’ standards bysample companies.

The remainder of this paper is organized as follows. Section II provides a review of the literature and develops the hypotheses. Section III describes our methodology and pre-sents and discusses empirical results and sensitivity tests, while Section IV provides con-cluding remarks.

II. LITERATURE REVIEW, HYPOTHESES DEVELOPMENT, AND OVERVIEWOF SECURITIES REGULATION

Literature Review and Hypotheses DevelopmentEinhorn (2005) develops a model which shows that a firm’s voluntary disclosure strat-

egy is crucially affected by the presence of its home country mandatory disclosure require-ments. Specifically, Einhorn (2005) shows that firms’ overall disclosures might be enhancedby limiting their discretion in mandatory reporting or by extending the scope of mandatorydisclosure requirements. She argues that a significant correlation exists between manda-tory and voluntary disclosures of a firm and uses a theoretical model to support her argu-ment. Einhorn (2005) analyzes how the incentives of firms to disclose information volun-tarily relate to various features of the mandatory disclosure regime. In a ‘‘flexible disclosureregime’’ and ‘‘no disclosure regime,’’ for example, there is more discretion in mandatory

reporting.3

This discretion enlarges the managers’ set of disclosure alternatives, enabling adisclosure strategy that minimizes the use of voluntary disclosure. Similarly, Einhorn (2005)compares the managers’ alternatives under the ‘‘strict disclosure regime’’ and reports thatthe overall disclosures of the firms in this category will be enhanced. Therefore, this modelpredicts that those USLAC, which have stricter requirements for mandatory disclosures intheir home country, will make more voluntary disclosures in the U.S. This leads to thefollowing hypothesis (stated in alternative form):

H1: For USLAC, a stricter mandatory disclosure environment in their home countrywill be associated with greater voluntary disclosures in the U.S.

3 Note that a ‘‘no disclosure regime’’ refers to no ‘‘mandatory disclosures.’’ In this regime, all disclosures aretechnically voluntary (since none is mandatory). Einhorn (2005) uses the ‘‘no disclosure’’ framework as a startingpoint for examining flexible disclosure regimes


4/27


A second purpose of this study is to examine the extent to which the voluntary disclo-sures of sample companies are convergent and the effect of domestic and global cultureon these disclosures. We use Gray’s (1988) theory of cultural influence on accounting dis-closures, which is in turn derived from Hofstede’s (1980) four cultural dimensions

(individualism-collectivism, masculinity-femininity, power distance, and uncertainty avoid-ance). Hofstede (1980) defines culture as ‘‘the collective programming of the mind’’ whichdistinguishes the members of one human group from another. Gray’s theory explains howculture affects the development of businesses and their institutions, including accountingsystems. We provide a brief overview of Hofstede’s cultural dimensions in the paragraphsbelow.

Individualism represents the degree of separateness within a society and a preferencefor a loosely knit social framework in society in which individuals are expected to takecare of themselves. In contrast, collectivism is associated with families and groups thatfoster secrecy and represents a preference for a tightly knit social framework. Individualisticsocieties are less likely to have developed large closely held companies and are likely to

be less dependent on banking relationships for capital. Therefore, it is expected that com-panies in an individualistic society are more likely to disclose higher levels of publicinvestor-oriented information.4

Power distance, another cultural dimension, denotes the dispersion of authority in asociety. In a society with a high power distance, for example, there is less dispersion of and less questioning of authority figures and vice-versa. Societies with high power distanceare, therefore, likely to have developed businesses and related institutions that discourageextensive sharing of information and vice-versa.5

As can be expected, there is a high level of correlation between individuality/mascu-linity, and power distance/uncertainty avoidance. Accordingly, this study will examine the

effect of the culture dimensions of individuality and power distance on the voluntary dis-closures of USLAC.Warner (2003) demonstrates that cultural differences have resulted in different man-

agement styles between Asian countries. A natural extension of this finding is that country-specific differences in management style results in country specific differences in voluntarydisclosure practices. Zarzeski (1996), however, argues domestic culture does impact vol-untary disclosure practices, but that a global culture may attenuate these effects whencompanies are listed on foreign markets.6 The present study extends the work of Warner(2003) and Zarzeski (1996) by examining the impact of culture on voluntary disclosurepractices of Asian companies listed on foreign markets.

4 Masculinity is a cultural tendency toward assertiveness, achievement, heroism, and material success, as opposedto feminism, which is associated with relationships, modesty, caring for the weak, and the quality of life.Countries that are more masculine are likely to be growth-oriented, economically and otherwise. Therefore, tocompete cost effectively in the business sector, companies in masculine societies are more likely to disclosehigher levels of information. Individuality and masculinity are typically highly correlated cultural dimensions.In the current study, their correlation was .746.

5 The cultural dimension of uncertainty avoidance signifies the degree to which a society can accept uncertaintyand ambiguity. Companies from strong uncertainty avoidance countries are, therefore, more likely to discloseless information publicly because more certain relationships exist. On the other hand, societies with weak uncertainty avoidance are more likely to have developed business relationships and institutions. Such societiesare more likely to have companies with widely held ownership and active capital markets. Power distance anduncertainty avoidance were correlated at .852 in the current study.

6 Two other studies indirectly relate to this issue. Cahan et al. (2005) suggest that the level of voluntary disclosures

is positively related to the extent of global operations. In addition, Khanna et al. (2004) finds that greater U.S.market interaction is associated with greater conformity with U.S. disclosure practices.


5/27


Specifically, we compare the voluntary disclosures by all USLAC in the U.S., measuredby Botosan (1997), to examine the effect of home culture and ‘‘global’’ culture. Resultsshowing a borrowing of ‘‘global’’ culture would provide evidence on convergence and vice-versa. Based on the correlations previously discussed, we use the two cultural dimensions

of individualism and power distance. This leads to the following two hypotheses (stated inalternative form):

H2: USLAC from individualistic countries are likely to provide more voluntary dis-closures, in the U.S., than USLAC from collectivist countries.

H3: USLAC from low power distance countries are likely to provide more voluntarydisclosures, in the U.S., than USLAC from high power distance countries.

The third objective of the research is to investigate the voluntary use of ‘‘international’’standards instead of national standards. Tarca (2004) examines the reporting practices of

foreign listed and domestic-only listed companies from the U.K., France, Germany, Japan,and Australia to determine the extent to which companies voluntarily use ‘‘international’’standards instead of national standards. As mentioned previously, ‘‘international’’ standardsare defined as U.S. GAAP or International Accounting Standards (IAS—now IFRS). Tarca(2004) tests for a preference for either set of standards, and also considers the relationshipof choice of regime with firm attributes. Tarca (2004) reports that companies using ‘‘inter-national’’ standards are likely to be larger and have more foreign revenue. Therefore, it ispredicted that the use of ‘‘international’’ standards among USLAC will be positively as-sociated with firm size and proportion of foreign revenue. This leads to the following twohypotheses (stated in alternative form):

H4: Larger USLAC are more likely to use ‘‘international’’ standards.

H5: USLAC with a greater proportion of foreign revenue are more likely to use‘‘international’’ standards.

Overview of Securities Regulation in the United States

Under the Securities Exchange Act of 1934, a foreign issuer must periodically fileinformation with the Securities and Exchange Commission (SEC) once it has listed itssecurities on a national securities exchange or NASDAQ, or once it has registered a secu-rities offering under the Securities Act of 1933.7 Foreign issuers subject to the SEC filingrequirement must file a Form 20-F annual report electronically. The financial statements in

Form 20-F need not be prepared in accordance with U.S. GAAP if they are presented inaccordance with another comprehensive body of GAAP and are accompanied by quanti-tative reconciliation of the materially different items in the financial statements to U.S.GAAP.8

With regard to the timing of these disclosures, foreign issuers in the U.S. must fileannual reports on Form 20-F within six months of fiscal year-end, and interim reports if and when they are released in their home country. However, it has been documented in theliterature that the American governance rules affecting U.S.-listed foreign firms are much

7 International Reporting and Disclosure Issues in the Division of Corporation Finance (SEC 2004).8 However, according to a recent SEC ruling, foreign companies preparing their financial statements in accordance

with IFRS as issued by the IASB for financial years ending after November 15, 2007 need not provide recon-ciliation to U.S. GAAP (Source: http://www.sec.gov/rules/final/2007/33-8879.pdf).

http://www.sec.gov/rules/final/2007/33-8879.pdfhttp://www.sec.gov/rules/final/2007/33-8879.pdf


6/27


stricter in writing than in practice (Siegel 2005). Siegel (2005) examines the SEC’s 70-yearenforcement record against all U.S.-listed foreign firms and reports that the SEC has rarelyenforced or been able to enforce U.S. securities laws against any of those firms. Siegel(2005) indicates, for example, that prosecuting international firms is difficult due to inter-

national obstacles and hence, political pressure on non-U.S. firms is more limited.

Overview of Securities Regulation in Asia

The securities regulation in Asia is fairly similar across all Asian countries being ex-amined, in terms of number of reports required. For example, the domestic stock exchangesin all sample countries require, from the listed companies, annual reports, as well as interimhalf-yearly statements. Seven of our nine sample countries (all except Hong Kong and thePhilippines) also require quarterly financial statements. All Asian countries’ stock exchangeshave requirements providing a basic framework and minimum standard for many financialdisclosures. However, in some countries (Singapore and Japan are examples) the securitiesmarket regulator provides a list of disclosures over and above the basic information thatmust be provided in the annual report.

In all sample countries, the structure of the securities market is similar in that there isa securities market regulator, which has the authority for regulation, development, andsupervision of securities markets. Seven of the nine sample countries have more than onestock exchange while South Korea and Taiwan are countries that have only one stock ex-change each. This study has reviewed the disclosure regulations of the primary stock exchange in each country to ensure that none of the required disclosures in their homecountries is present in the Botosan (1997) disclosure index. See Table 1 for general infor-mation regarding the primary stock exchange of each of the Asian countries used in thecurrent study.

III. RESEARCH METHODOLOGYSample Selection and Description

The sample for this study consists of all USLAC in the years 2003 and 2004. Thissample period was selected because the SEC required electronic submission of Form 20-Fannual reports by foreign companies after May 2002. This leads to 96 companies from nineAsian countries (see Table 2). However, 12 of these companies are listed on over-the-counter (OTC) debt and stock exchanges.9 There are significant differences in the disclosurerules for companies listed on OTC exchanges versus those listed on national stock ex-changes such as the New York Stock Exchange (NYSE) and NASDAQ. Therefore, this

paper focuses on the 84 Asian companies (from nine countries) listed on the NYSE/ NASDAQ. A combined two-year cross-sectional analysis is conducted to test the hypoth-eses. In addition, each year (2003 and 2004) is examined separately.

Dependent Variables

An Ordinary Least Squares (OLS) regression model and a binomial logistic regressionmodel are used to test the hypotheses. The dependent variable in the OLS regression modelis a score based on total voluntary disclosures provided by the sample firms in their Form20-F annual report. Voluntary disclosures have been defined as those disclosures that exceedrequirements and represent free choices on the part of company’s management to provide

9 The number of USLAC on the OTC is 12 out of 96 companies (12.5 percent). The number of all foreign OTCcompanies on the U.S. stock exchanges is 450 out of 1,240 companies (36.29 percent).


7/27


TABLE 1

Primary Stock Exchanges in Asiaa,b,c

CountryPrimary Stock

Exchange

Number of Listed

Companies

MarketCapitalization(in millions of U.S. Dollars)

Gross

DomesticProduct (inmillions of

U.S. Dollars)

1. China Shanghai Stock Exchange 947 $314,304.32 $1,649,329.00

2. Hong Kong Stock Exchange of HongKong

906 851,083.79 163,005.00

3. India Bombay Stock Exchange 9359 383,360.00 691,876.00

4. Indonesia Jakarta Stock Exchange 334 355.31 257,641.00

5. Japan Tokyo Stock Exchange 2276 3,172,640.78 4,623,398.00

6. Philippines Philippine Stock

Exchange

235 85,683.68 86,429.00

7. Singapore Singapore Exchange 638 90,819.77 106,818.00

8. South Korea Korea Stock Exchange 684 360,263.47 679,674.00

9. Taiwan Taiwan Stock Exchange 700 419,186.74 305,200.00

a This table provides the market capitalization of the primary stock exchange from each sample country inmillions of U.S. dollars as of December 31, 2004.

b The exchange rates are taken from the Federal Reserve website.c Data is available at the respective stock exchanges’ websites.

TABLE 2

Number of U.S.-Listed Asian Companies (USLAC)a,b

Country Number of Companies

1. China 13

2. Hong Kong 8

3. India 10

4. Indonesia 2

5. Japan 26

6. Philippines 2

7. Singapore 6

8. South Korea 10

9. Taiwan 7

Total 84

a This table provides the number of companies from each Asian country listed in the United States excluding theOver-the-Counter (OTC) companies.

b This table does not include those USLAC which are listed on the OTC exchanges. The following is thenumber of OTC companies from each Asian country: China: 0, Hong Kong: 4, India: 0, Indonesia: 0, Japan: 4,Korea: 2, Philippines: 2, Singapore: 0, Taiwan: 0.

financial information deemed relevant to the decision needs of users (Radebaugh and Gray1997). Other disclosure studies have used a number of different approaches to measure the

voluntary disclosures provided by firms (see Frost and Pownall [1994], Meek et al. [1995],Frost and Kinney [1996], and Khanna et al. [2005]). This study uses the disclosure index


8/27


TABLE 3

Disclosure Componentsa,b

Variables

Loadings

Current Disclosures Prospective Disclosures

Background Information .639 .445

Summary of Historical Information .001 .476

NonFinancial Information .854 .295

Prospective Information .141 .795

Management Discussion & Analysis .501 .343

a This table provides the factor loadings of the five components of Voluntary Disclosures provided by USLAC inthe U.S.

b Based on Varimax rotation.Variable Definitions:

Current Disclosures current disclosures mainly about background information, nonfinancialinformation and management discussion and analysis;

Prospective Disclosures prospective disclosures mainly about future forecasts and prospects; Background Information disclosure of background information;

Summary of Historical Information disclosure of historical information; NonFinancial Information disclosure of nonfinancial information;

Prospective Information disclosure of prospective information; and Management Discussion & Analysis disclosure of management discussion and analysis information.

developed in Botosan (1997). Botosan indicates that the purpose of this instrument is toprovide a cross-sectional ranking of voluntary disclosures provided by firms in their annualreports. Therefore, the Botosan instrument appears to be well suited for the current research.

The major components of the Botosan (1997) disclosure index are classified into the

following five categories: background information, summary of historical results, key non-financial statistics, projected information, and management discussion and analysis. Inves-tors and financial analysts have identified these categories as useful in investment decisionmaking. Each category includes specific disclosure questions that represent management’svoluntary disclosure of information for that category. Therefore, a particular questionis awarded points on the basis of the specific voluntary disclosures. The sample firms areawarded scores on the basis of quantity as well as quality of their voluntary disclosures.

In addition to total voluntary disclosures, this study factor analyzed its five categoriesand arrived at two reasonably distinct factors (similar to Cahan et al. 2005). Table 3 providesthe factor loadings of the five components of voluntary disclosures. One factor representsbackground information, nonfinancial disclosures, and management discussion and analysis,

while the other factor represents projected information. Due to the nature of disclosurescovered by each of the two factors, this study characterizes the background information,nonfinancial information, and MD&A factor as a current disclosure measure, and the projected information factor as a prospective disclosure measure.10 Historical results did notload heavily on either factor, so that item is not treated as a separate measure. This studyalso estimated the OLS regression models using these new measures (Current Disclosureand Prospective Disclosure) as alternative dependent variables. Similar to Botosan

10 There is a slight difference in these results and the results of factor analysis in Cahan et al. (2005). In Cahanet al. (2005), Current Disclosure loads heavily only on background and nonfinancial disclosures and Prospective

Disclosure loads heavily on projected information and MD&A. However, it can be argued that MD&A is asmuch a part of current disclosures as it is a part of prospective disclosures, because these disclosures primarilyinclude the management explanations for changes in current financial items


9/27


(1997) and Cahan et al. (2005), this study ranks the disclosure scores (including the factorscores for Current Disclosure and Prospective Disclosure) to convert the ordinal scores intolevels for ease of interpretation. All of the primary statistical analyses in this study arebased on these ranks.

The dependent variable in the binomial logistic regression model is a dichotomousvariable ( International Standard ) indicating the voluntary use of ‘‘international standards’’(1 for adoption and 0 otherwise). Tarca (2004) defines ‘‘international’’ standards as U.S.GAAP or IAS. Similar to Tarca (2004), this study identifies those companies as users of ‘‘international’’ standards that have stated in their accounting policy note or audit reportthat that they used U.S. GAAP or IAS to prepare their consolidated financial statements,instead of national GAAP. Companies that provided a reconciliation statement to U.S.GAAP are not classified as a user of ‘‘international’’ standards.

Independent Variables

A measure of the strictness of the mandatory disclosure environment is needed in orderto examine the issue of how home country mandatory disclosures affect the voluntarydisclosures, in the U.S., of Asian companies. This study measures the strictness of the man-datory disclosure environment ( Disclosure Strictness) by using experts/raters in the area of international accounting research (specifically the Asian-Pacific region). These experts wereasked to rank the nine sample countries/regions based on the strictness of their domesticmandatory disclosure environment.11 An instrument (see Appendix) was sent to the expertsrequesting them to rank the sample countries from one (most strict) to nine (least strict).

Nine experts were selected after a careful consideration of their publication record.In the selection of experts, primary consideration was given to the proportion of theirpublications in accounting research, specifically in the Asian continent. Seven of the nine

experts have over 20 publications in the area of International accounting research, with asignificant percentage of those studies investigating Asian accounting issues. Three of theexperts are country representatives for China, Hong Kong, and Singapore in the AmericanAccounting Association International Accounting Section. One expert is the editor of Jour-nal of International Accounting Research. The instrument was mailed electronically to theseexperts and their responses were tested for consistency using Kendall’s W statistic. Table4 provides the determination of the Disclosure Strictness variable by using their ratings.To determine whether there was any consensus in the mean rankings by experts, this studycomputed Kendall’s W statistic12 and the average Spearman correlation on the rankings of all possible pairs of judges. Kendall’s W statistic is 0.655 and statistically significant (at

less than the 0.001 level). The average Spearman correlation computed on the rankings of all possible pairs of judges is 0.606. Therefore, it is reasonable to conclude that there is aconsensus among the experts’ ratings of sample countries.13

11 Hong Kong is governed as a Special Administrative Region (SAR) of China. Accordingly, we use the term‘‘countries/regions’’ due to the governing status of Hong Kong.

12 Kendall’s W statistic determines the extent of agreement in ranking between two or more experts, exceedingthat which would be expected by chance (Pett 1997).

13 To further explore the consensus of expert rankings, we performed an analysis to ensure that none of the expertscontributed excessively to a lower Kendall’s W and Spearman correlation. Specifically, we recomputed Kendall’sW and the Spearman correlation by leaving out one expert at a time. The highest consensus measure was reachedby leaving out Expert 6 (Kendall’s W .721 and Spearman correlation .681). Further, we recalculated themean rankings by leaving out Expert 6, but this measure was nearly perfectly correlated (.9907) with our originaldetermination of mean rankings. We also reexamined the regression models using mean rankings (without Expert6), but not surprisingly the results were virtually identical. Accordingly, the results reported in the manuscriptare based on mean rankings from all eight experts


10/27


TABLE 4

Determination of Disclosure Strictnessa,b

Country Mean Ranking

1. China 7.382. Hong Kong 2.38

3. India 5.63

4. Indonesia 7.88

5. Japan 3.50

6. Philippines 6.63

7. Singapore 1.50

8. South Korea 4.38

9. Taiwan 5.75

Kendall’s W (p-value) Experts

0.655(0.000)

8

a This table provides the determination of the Disclosure Strictness variable on the basis of mean rankingsprovided by experts (the lower the mean ranking, the stricter the disclosure requirements).

b The instrument was emailed to nine experts, and eight experts provided their responses.

Two cultural dimensions, Individualism and Power Distance, are measured as contin-uous variables to examine the effect of ‘‘global’’ culture on the USLAC. These measureswere developed by Hofstede (1980) in a multidimensional scaling of work-related surveys

from over 160,000 IBM employees across 64 countries. See Table 5 for a list of scores onthese variables.14

Prior research has identified several variables that affect the voluntary disclosures pro-vided by firms. These variables include firm size (Zarzeski 1996; Hope 2003; Tarca 2004),percentage of foreign sales (Zarzeski 1996; Tarca 2004; Cahan et al. 2005), and leverageratio (Zarzeski 1996; Tarca 2004). Hope (2003) documents that there is a positive relationbetween size and voluntary disclosures of a firm while Tarca (2004) finds that the firm sizeis positively associated with the voluntary use of ‘‘international’’ standards. In addition,

14 Recall that in our study, we asked experts to rank nine Asian countries according to the strictness of the man-

datory reporting requirements of that country. We did not, however, ask the experts to classify the nine countriesaccording to the individualistic and power distance cultural variables. Accordingly, we are unable to directlyreconcile our expert rankings with the cultural rankings by Hofstede (1983). However, based on theory and priorresearch, we understand that the cultural constructs of individualism and power distance, as developed byHofstede, may be related to the cultural influences that result in the strictness of mandatory reporting require-ments. Specifically, we expect that countries with individualistic cultures, because of the separateness within thesociety, will be more likely to disclose more investor-oriented information. Alternatively, we expect that countriesidentified with high degrees of power distance will be less likely to disclose investor-oriented information becauseof a greater degree of dispersion of authority and the resulting unwillingness to share information. Accordingly,we would expect our ‘‘strictness’’ measure, as reported by experts, would be positively related to individualismand negatively correlated with power distance. An analysis supports this expectation. When the expert meanrankings are correlated with the Hofstede (1983) individualism rankings for the countries in question, we findthat more individualistic countries (as identified by Hofstede) are considered to have more strict reportingrequirements by the experts who participated in the present study (Spearman rho .310). Alternatively, countries

with high power distance were considered to have less strict mandatory reporting requirements by experts(Spearman’s rho .476).


11/27


TABLE 5

Culture Scores for Asian Countriesa

Countries Power Distance Individualism / Collectivism

1. China 80 202. Hong Kong 68 25

3. India 77 48

4. Indonesia 78 14

5. Japan 54 46

6. Philippines 94 32

7. Singapore 74 20

8. South Korea 60 18

9. Taiwan 58 17

a These scores were developed by Hofstede (1980) in a multidimensional scaling of work-related surveys of

160,000 IBM employees over 64 countries. A higher value indicates more of that particular cultural trait.

Hossain et al. (1995) argue that firm size is an effective proxy for omitted variables. There-fore, this study uses Firm Size as an independent variable, measured as the natural log of the firm’s total assets in millions of U.S. dollars.

Zarzeski (1996) and Cahan et al. (2005) report that companies with more foreign saleswill provide additional disclosures because they are more internationally dependent andhave higher agency costs. Tarca (2004) documents that companies with higher foreign salesare more likely to adopt ‘‘international’’ standards instead of national standards. Therefore,this study uses percentage of foreign sales (Foreign Sales%) as an independent variable

(foreign sales divided by total sales), measured in millions of U.S. dollars.Zarzeski (1996) and Tarca (2004) argue that companies with higher debt ratios are

likely to share more information with their creditors because they usually develop bank-ing relationships and interlocking corporate ownerships as alternate capital sources to publicownership. Conversely, companies with lower debt ratios have a higher percentage of stock ownership, which could encourage investor demand for information (i.e., companieswith lower debt ratios are expected to have higher levels of investor disclosures). Therefore,the present study uses Debt Ratio as an independent variable (total debt divided by totalassets).

Descriptive and Preliminary AnalysesTables 6 through 10 provide descriptive statistics and the results of preliminary anal-

yses. Table 6 shows the descriptive statistics for Voluntary Disclosures and the independentvariables. The mean for the overall voluntary disclosure score is 48 with a range from 32to 70.15 Of the five components of Voluntary Disclosures, background information andMD&A disclosures make up 56 percent of the mean of the total score.16 There is also large

15 The mean score in this study is well above Botosan’s (1997) score of 30 and slightly above the Cahan et al.(2005) score of 44.431. However, the range of Voluntary Disclosure in Cahan et al. (2005) is wider than therange in this paper (21 to 85). There could be two reasons for these differences. First, the sample in this studyincludes nonfinancial firms from a wide range of industries, whereas Botosan’s (1997) sample is from a narrowgroup of industrial firms. Second, the Cahan et al. (2005) sample has all Fortune 500 firms whereas this study’s

sample firms include all USLAC.16 Similar to Cahan et al. (2005).


12/27


TABLE 6

Descriptive Statisticsa

Variables Minimum Maximum Mean Standard Deviation

Voluntary Disclosures

Background Information 5 16 11 2.233

Summary of Historical Information 2 4 3 0.499

NonFinancial Information 4 18 9 4.060

Prospective Information 0 18 9 2.545

Management Discussion & Analysis 8 24 16 4.036

Total Voluntary Disclosures 32 70 48 8.220

Other Variables

International Standards 0 1 0.79 0.407

Power Distance 54 94 66 11.257 Individualism 14 48 32 13.379

Foreign Sales% 0 99.64 32.95 33.874

Firm Size 0.02 919.74 37.10 102.09

Debt Ratio 0.01 0.97 0.45 0.252

a This table provides the descriptive statistics for the dependent and independent variables that are used in thisresearch.

Variable Definitions:Voluntary Disclosures total disclosure score from the Form 20-F annual report as per Botosan

(1997) disclosure index; Background Information disclosure of background information;

Summary of Historical Information disclosure of historical information; NonFinancial Information disclosure of nonfinancial information;

Prospective Information disclosure of prospective information; Management Discussion & Analysis disclosure of management discussion and analysis;

International Standards dichotomous variable; (1 use of U.S. GAAP or IFRS, 0 otherwise);Power Distance sample country’s power distance score from Hofstede (1980);

Individualism sample country’s individualism score from Hofstede (1980);Foreign Sales% foreign sales divided by total sales as reported in the Form 20-F annual

report;Firm Size natural log of total assets of the firms (in billions of U.S. Dollars); and

Debt Ratio total debt divided by total assets (in millions of U.S. Dollars).

variation in the two cultural variables. The score on Power Distance ranges from 54 to 94with a mean of 66 and the score on Individualism varies from 14 to 48 with a mean of 32.

Table 7 gives a breakdown of the disclosure scores (mean and median scores) byindustry. Approximately 38 percent of the sample firms are from the Manufacturing sector,while 25 percent are from the Transportation and Communication sector. The highest meanand median Voluntary Disclosures score was for the Transportation and Communicationindustry, and the lowest was for the Real Estate, Insurance, and Investment industry. Similarto Cahan et al. (2005), nonfinancial disclosures and MD&A disclosures are the componentsof Voluntary Disclosures exhibiting the most variation among industries.

Table 8 provides a distribution of the disclosure scores (mean and median scores) bycountry. Of these 168 observations from nine countries, 31 percent are Japanese firms, 15.5percent are Chinese firms 12 percent of the firms are from India and another 12 percent


13/27

J o ur n a l o f I n t er n

a t i o n a l A c c o u n t i n gR e s e ar c h , V o l u m e 7 , N o .1 ,2 0 0 8

TABLE 7

Industry Disclosure Statisticsa

Industry (SIC Codes) n

VoluntaryDisclosures

Mean(Median)

BackgroundInformation

Mean(Median)

Summary of Historical

InformationMean

(Median)

NonFInfo

M(M

Mining and Construction(1221, 1311)

8 50(50)

12(11)

3(3)

Manufacturing(3570, 3577, 3651, 3821,

3320, 3600, 3944, 3695)

64 50(48)

11(11)

3(3)

Transportation, Communicationand Utilities(4512, 3663, 4812, 4011,

4813)

42 52(53)

11(11)

3(3)

Trading-Wholesale and Retail(5050, 2340)

4 45(45)

10(10)

3(3)

Real Estate, Insurance andInvestment(6029, 6331, 6311, 6163,

6211)

22 39(40)

11(10)

3(3)

Service Industries—Recreation(4841, 2711, 7372)

8 41(39)

10(10)

3(3)

Service Industries—Others(4911, 7389, 7371, 7370) 20 47(48) 12(12) 2(2)

a This table provides the mean and median disclosure statistics for the sample firms in different industries.


14/27

J o ur n a l o f I n t er n a t i o n a l A c c o u n t i n gR e s e ar c h , V o l u m e 7 , N o .1 ,2 0 0 8

TABLE 8

Country Disclosure Statistics with Respect to Disclosure Stri

Country(Mean Ranking for Disclosure Strictness) n


Mean(Median)


Mean(Median)

Summary of Historical

InformationMean

(Median)

NonFinInform

Mea(Med

Singapore(1.50)

12 46(43)

10(9)

3(3)

8(7

Hong Kong(2.38)

16 45(46)

10(10)

3(3)

9(9

Japan(3.50)

52 46(46)

11(11)

3(3)

6(4

South Korea(4.38)

20 50(50)

12(12)

3(3)

12(12

India(5.63)

20 45(46)

12(11)

3(3)

6(6

Taiwan(5.75)

14 52(52)

11(11)

3(3)

8(6

Philippines(6.63)

4 47(46)

8(8)

3(3)

9(9

China(7.38)

26 54(56)

12(11)

3(3)

13(14

Indonesia

(7.88)

4 48

(48)

11

(11)

3

(3)

11

(11a This table provides the mean and median disclosure scores for companies with respect to the strictness in their man

(the lower the mean ranking, the stricter the disclosure regime).


15/27


are from South Korea. The median Voluntary Disclosure scores are highest for Chinesefirms (56) and lowest for Singapore firms (43).

Table 9 provides the results of t-tests comparing the characteristics of companies whenexamining Disclosure Strictness in a dichotomous sense (strict mandatory and nonstrict

mandatory regimes, respectively). Companies in the nonstrict mandatory group (mean 50.103) provide more overall Voluntary Disclosures as compared to companies in thestrict mandatory group (mean 46.760). Among the five categories of Voluntary Disclo-sures, companies in the strict mandatory group provide significantly more disclosures thanthe nonstrict group in only the projected information category. In the nonfinancial andMD&A disclosures categories, nonstrict mandatory regime companies provide significantlymore disclosures.

There are significant differences between the two groups on both cultural dimensions.Companies in the strict mandatory group have low Power Distance. In addition, the do-mestic culture of sample companies in the strict mandatory group is dominated by individ-uality. Recall that Power Distance denotes the dispersion of authority in a society, and Gray

(1988) suggests that the countries with low Power Distance are likely to provide moreVoluntary Disclosures. However, the results in Table 9 suggest that companies whose do-mestic culture is dominated by low Power Distance are providing fewer Voluntary Disclo-sures.17 Individualism is defined in Gray (1988) as the degree of separateness within asociety, and it is expected that companies in an individualistic society are more likely todisclose higher levels of public investor-oriented information. However, the results in Table9 suggest otherwise, indicating that the companies whose domestic culture is dominated byindividuality are providing fewer voluntary disclosures.

Table 10 provides the number of companies from each country that are using domesticGAAP, U.S. GAAP, and IFRS in their Form 20-F annual report, and their mean and median

Voluntary Disclosure scores respectively. In 35 out of 168 observations, firms are usingtheir domestic standards to prepare their financial statements in the Form 20-F annual report.Of the remaining 133 observations, 72 percent are using U.S. GAAP, and 28 percent areusing IFRS. In none of the sample countries are companies using all three types of standards(domestic GAAP, U.S. GAAP, and IFRS).

Three of the nine Asian countries have some companies using IFRS. From China, fourobservations are under domestic GAAP, and the remaining 22 are under IFRS (disclosurescores are higher for companies using IFRS). In Hong Kong, two are using U.S. GAAPwith the remaining 14 using IFRS (the disclosure scores are higher for companies usingIFRS). In India, two are using IFRS, and the remaining 18 are using U.S. GAAP. Threeother Asian countries have all companies using the same type of GAAP. In Indonesia, all

sample companies are using domestic GAAP, while in Japan and the Philippines, all com-panies are using U.S. GAAP.

The remaining three Asian countries have companies using either U.S. GAAP or do-mestic GAAP. In Singapore, three observations are under domestic GAAP and the remain-ing nine are under U.S. GAAP.18 Companies using U.S. GAAP are providing more overallvoluntary disclosures. In South Korea, 12 use domestic GAAP and eight use U.S. GAAP.South Korean companies using domestic GAAP provide more voluntary disclosures than

17 Zarzeski (1996) reports that local companies, but not international companies, provide disclosures commensuratewith the secretiveness of their local culture. Therefore, one plausible explanation for the low power distancecountries providing fewer voluntary disclosures is that when companies from these countries are listed on global

stock exchanges, the effect of their domestic culture is reduced.18 One company used domestic GAAP in 2003 and used U.S. GAAP in 2004.


16/27


TABLE 9

Comparison of Means for Variables with Respect to Disclosure Strictnessa,b

Variable

Disclosure

Strictness n Mean t-value

p-value

(two-tailed)Voluntary Disclosures NonStrict 68 50.103 2.633 0.009

Strict 100 46.760

Background Information NonStrict 68 11.176 0.558 0.577Strict 100 10.980

Summary of Historical Information NonStrict 68 2.809 0.366 0.715Strict 100 2.780

NonFinancial Information NonStrict 68 9.823 3.022 0.003Strict 100 7.940

Projected Information NonStrict 68 8.823 2.743 0.007Strict 100 9.900

Management Discussion & Analysis NonStrict 68 17.470 3.785 0.000Strict 100 15.160

Current Disclosures NonStrict 68 0.328 3.631 0.000Strict 100 0.222

Prospective Disclosures NonStrict 68 0.181 1.952 0.053Strict 100 0.123

Power Distance NonStrict 68 75.294 11.815 0.000Strict 100 59.840

Individualism NonStrict 68 27.970 2.891 0.004Strict 100 33.920

Foreign Sales% NonStrict 68 32.039 0.287 0.775Strict 100 33.570

Firm Size NonStrict 68 8.064 2.843 0.005Strict 100 9.024

Debt Ratio NonStrict 68 0.373 3.434 0.001Strict 100 0.506

a This table provides the comparison of means for dependent and independent variables of companies whichhave strict and non–strict mandatory disclosure regimes.

b Disclosure Strictness was split into two groups—Strict and NonStrict—on the basis of mean rankings providedby experts. The countries with a mean ranking below the midpoint of the mean rankings were classified asStrict and countries with a mean ranking above the midpoint were classified as NonStrict.

those using U.S. GAAP. In Taiwan, 12 are using domestic GAAP and the remaining twoare using U.S. GAAP.

Regression Results

Tables 11 to 13 provide the results of the OLS regressions for the overall sample andTable 14 provides the results of binomial logistic regression. To test H1, H2, and H3, thefollowing OLS regression was run on three alternate dependent variables (Voluntary Dis-closures, Current Disclosures, and Prospective Disclosures):

Voluntary Disclosures Disclosure Strictness Power Distancei 0 1 2 c

Individualism Foreign Sales%3 c 4 i

19 Firm Size Debt Ratio .

5 i 6 i


17/27

J o ur n a l o f I n t er n

a t i o n a l A c c o u n t i n gR e s e ar c h , V o l u m e 7 , N o .1 ,2 0 0 8

TABLE 10

Voluntary Disclosures of Companies Using Domestic GAAP, U.S. GA

Country

(n) n


Mean

(Median)


Mean

(Median)

Summary of Historical Information

Mean

(Median)

NonFinancialInformation

Mean

(Median)

Domestic GAAP

China 4 50(50)

10(10)

3(3)

13(12)

Indonesia 4 48(48)

11(11)

3(3)

11(11)

Singapore 3 44(42)

9(8)

3(3)

5(4)

South Korea 12 53(54)

12(12)

3(3)

12(13)

Taiwan 12 50(49)

11(11)

3(3)

7(6)

U.S. GAAP

Hong Kong 2 38(38)

11(11)

3(3)

4(4)

India 18 45(46)

12(12)

3(3)

6(5)

Japan 52 46(46)

11(11)

3(3)

6(4)

Philippines 4 47(46)

8(8)

3(3)

9(9)

Singapore 9 46(43)

10(9)

3(3)

9(8)

South Korea 8 46(45)

12(12)

3(3)

12(12)

Taiwan 2 64(64)

15(15)

3(3)

16(16)


18/27

J o ur n a l o f I n t er n a t i o n a l A c c o u n t i n gR e s e ar c h , V o l u m e 7 , N o .1 ,2 0 0 8

TABLE 10 (continued)

Country(n) n


Mean(Median)


Mean(Median)

Summary of Historical Information

Mean(Median)

NonFinancialInformation

Mean(Median)

IFRS

China 22 55(56)

12(12)

3(3)

14(14)

Hong Kong 14 46(46)

10(10)

3(3)

9(10)

India 2 45(45)

11(11)

3(3)

8(8)

a This table provides the mean and median voluntary disclosure scores for the companies from each country using do


19/27


TABLE 11

Regression Results for the Overall Sample

(Dependent Variable: Voluntary Disclosures)

Exp. Sign Std. Beta t p-valueIntercept 19.733 3.221 0.002

Disclosure Strictness 0.282 3.232 0.001

Power Distance 0.139 1.449 0.149

Individualism 0.240 3.137 0.002

Foreign Sales% 0.052 0.659 0.511

Firm Size 0.206 2.225 0.027

Debt Ratio 0.304 3.505 0.001

Adj R2 F p-value

Model Summary 0.192 7.615 0.000

TABLE 12


(Dependent Variable: Current Disclosures)

Exp. Sign Std. Beta t p-value

Intercept 62.032 2.427 0.016




Foreign Sales% 0.162 2.261 0.025Firm Size 0.226 2.671 0.008

Debt Ratio 0.309 3.886 0.000

Adj R2 F p-value

Model Summary 0.323 14.251 0.000

TABLE 13


(Dependent Variable: Prospective Disclosures)

Exp. Sign Std. Beta t p-valueIntercept 94.681 3.703 0.000




Foreign Sales% 0.369 4.996 0.000

Firm Size 0.081 0.932 0.353

Debt Ratio 0.137 1.677 0.095

Adj R2 F p-value

Model Summary 0.279 11.786 0.000


20/27


Regarding control variables, Firm Size is positive and statistically significant for bothregressions related to the Voluntary Disclosures and Current Disclosures variables. Thisfinding is consistent with prior research and indicates that the larger the firm, the greaterthe Voluntary Disclosures. The coefficient on Debt Ratio is negative and statistically sig-

nificant for both Voluntary Disclosures and Current Disclosures. This finding is also con-sistent with prior research and indicates that the higher the proportion of debt in the fi-nancing of a company, the lower the Voluntary Disclosures. The coefficient on ForeignSales% is not significant in the model for Voluntary Disclosures but is significant in theother models. It is positive and significant for Prospective Disclosures and is negative andsignificant for Current Disclosures. This result suggests that companies with a higher pro-portion of foreign sales are providing more voluntary disclosures in the projected infor-mation category and fewer voluntary disclosures in the background information, nonfinan-cial statistics, and MD&A disclosure categories.

The independent variable of primary interest for hypothesis one is Disclosure Strictness.This variable is constructed from the mean rankings of experts (with lower rankings indi-

cating a more strict mandatory regime and higher rankings indicating a less strict regime).Therefore, to reject the null form of the hypothesis, the coefficient on Disclosure Strictnessshould be negative and statistically significant. The coefficient on Disclosure Strictness ispositive in all regressions. This coefficient is positive and statistically significant for theVoluntary Disclosures and Current Disclosures regressions. This result is also supported inthe descriptive statistics. Therefore, H1 fails to be rejected. In fact, contrary to the Einhorn(2005) model, the results suggest that a nonstrict mandatory disclosure regime results ingreater voluntary disclosures (in the U.S.) by USLAC.

In order to reject the second hypothesis (H2), the coefficient on Individualism shouldbe positive and statistically significant. The results (Tables 11 through 13) show that this

coefficient is negative and significant for Voluntary Disclosures and Current Disclosuresand positive and marginally significant for Prospective Disclosures (after controlling forproportion of foreign sales, firm size, debt ratio, and domestic mandatory disclosure envi-ronment). Therefore, these results do not provide support to reject the second hypothesis(H2). Instead, the results suggest that companies from individualistic countries are likelyto provide fewer voluntary disclosures than companies from collectivist countries (at leastfor the Voluntary and Current Disclosure variables).

The third hypothesis (H3) predicts that the coefficient on Power Distance will be neg-ative and statistically significant. The results in Tables 11 through 13 show that after con-trolling for the aforementioned variables, this coefficient is negative and statistically sig-nificant only when Prospective Disclosures is the dependent variable. Therefore, this result

provides support for rejecting the third hypothesis (H3), at least for the Prospective Dis-closures variable.

The following binomial logistic regression equation was used to examine the effect of firm size and percentage of foreign sales on the choice of ‘‘international’’ standards byUSLAC in the U.S.:

Logit (International Standard ) Power Distance Individualismi 0 1 c 2 c

Foreign Sales% Firm Size3 i 4 i

20 Debt Ratio .

5 i

20 Subscripts: i firm-specific; c country-specific.


21/27


TABLE 14

Binomial Regression Results

(Dependent Variable: International Standard)

Exp. Sign Std. Beta t p-value

Intercept 26.404 5.130 0.024



Foreign Sales% 0.011 1.229 0.268

Firm Size 0.237 1.203 0.273

Debt Ratio 1.512 0.745 0.388

NagelkerkeR2

WaldStatistic p-value

Model Summary 0.704 49.383 0.000

The fourth hypothesis (H4) predicts that the coefficient on Foreign Sales% will bepositive and statistically significant. The results in Table 14 show that, after controlling forrelevant variables, this coefficient is not statistically significant. Therefore, this result failsto reject the fourth hypothesis (H4) and suggests that the USLAC choice to use ‘‘interna-

tional’’ standards was not affected by the proportion of foreign sales to their total sales.According to the fifth hypothesis (H5), the coefficient on Firm Size should be positiveand statistically significant. The results in Table 14 show that, after controlling for relevantvariables, this coefficient is also not statistically significant. Therefore, this result fails toreject hypothesis five (H5) and suggests that the size of USLAC did not affect their choiceto use ‘‘international’’ standards.

Sensitivity Analyses

Various sensitivity tests were conducted to determine whether the results are beingdriven by other extraneous factors. First, the OLS regression tests were repeated by usingan alternative measure of Disclosure Strictness. Adhikari and Tondkar (1992) report that incountries with higher domestic market capitalization (adjusted for size of economy), stock exchanges are likely to have more rigorous disclosure requirements. Therefore, we scaledthe market capitalization of all sample countries by their Gross Domestic Product (MarketCapitalization/GDP). We used this scaled measure to rank countries based on DisclosureStrictness. The Adhikari and Tondkar (1992) Disclosure Strictness measure was highlycorrelated with our expert measure (r 0.533). Furthermore, the regression results weresimilar (both in direction and magnitude) under both the expert measure and the Adhikariand Tondkar (1992) measure of Disclosure Strictness.

Second, the OLS regression tests for the overall sample were repeated with indicatorvariables for five industries (mining and construction, manufacturing, transportation, com-

munication and utilities, real estate, insurance and investment [REIT], and Service). Theresults (not tabulated) did not change from those previously reported (i.e., the coefficienton Disclosure Strictness was negative and statistically significant, even after controlling forthe industry effects). In addition, the results of the two hypotheses related to culture didnot change after controlling for the industry effects. Of the five industry variables, thecoefficients on mining and construction and manufacturing were positive and statistically


22/27


significant (when Current Disclosures was used as a dependent variable). This result sug-gests that the sample companies from mining and construction, and manufacturing indus-tries provided significantly higher current voluntary disclosures than companies in the otherindustries.

To investigate the effect of the culture variables on the voluntary disclosures of samplecompanies, without considering the strictness of their domestic mandatory disclosure re-gimes, the regression equations were run without the Disclosure Strictness variable. Theresults (not tabulated) of H2 and H3 remain generally the same as discussed previously.

To identify whether the stock exchange listing (NYSE or NASDAQ) affects the choiceto use ‘‘international’’ standards, an indicator variable for NYSE / NASDAQ listing was usedin the binomial logistic regression. However, the results (not tabulated) did not change andthe coefficients on Foreign Sales% and Firm Size were not statistically significant. In ad-dition, the coefficient on the indicator variable for stock exchange listing was not significant.

Lastly, Table 10 shows that there is a difference among the voluntary disclosure scoresof companies which are using domestic GAAP, IFRS, and U.S. GAAP. Therefore, an al-

ternative version of the logistic regression model was run with the dependent variablehaving three levels (instead of two).21 Only three independent variables were used (FirmSize, Foreign Sales%, and Debt Ratio). The results (not tabulated) indicate that the com-panies using U.S. GAAP are larger in size (statistically significant at less than the 0.05level) than the companies using domestic GAAP. In addition, the results suggest that com-panies using U.S. GAAP have a larger proportion of foreign sales than companies usingIFRS (statistically significant at less than the 0.05 level). None of the other coefficients wassignificant in this regression.

Overall, the regression results were consistent for the Voluntary Disclosures (overallVoluntary Disclosures) and Current Disclosures (factor score for background information,

nonfinancial information, and MD&A disclosures) variables. However, in some cases, theresults on the Prospective Disclosures (factor score for projected information category)variable were different from the results for Voluntary Disclosures and Current Disclosures.These differences on the Prospective Disclosures variable should not affect the conclusionsof this study, as this variable is a factor score for one of the five categories of the overallVoluntary Disclosures and accounts for only 19 percent of the overall disclosures.

IV. CONCLUSIONIn this study, we examined three issues: (1) the effect of strictness of mandatory dis-

closure regime on the voluntary disclosures of USLAC; (2) the effect of domestic and‘‘global’’ culture on those disclosures; and (3) the effect of firm size and proportion of foreign sales on the voluntary use of ‘‘international’’ standards.

These issues were examined using a sample of 84 companies from nine Asian countriesover a period of two years. The results, regarding the first issue, indicate that USLAC withless strict mandatory disclosure regimes are likely to provide more voluntary disclosures inthe U.S. than are the companies with stricter mandatory disclosure regimes. This result isstatistically significant and robust across several model specifications. Therefore, these re-sults do not follow the predictions from the Einhorn (2005) model.22 One possible reasonfor this finding is that the companies that have a relatively less strict mandatory disclosureregime in their home country will try to provide the maximum disclosures to meet the

21 A multinomial logistic regression was run instead of a binomial logistic regression.22 It is important to note that the Einhorn (2005) model was not developed specifically for USLAC.


23/27


stringent disclosure expectations in the U.S. Research has shown that the disclosure regu-lations for domestic companies in the U.S. are comparable to the most stringent guidelinesanywhere in the world (e.g., Frost and Pownall 1994 and Frost and Kinney 1996). On theother hand, the companies which have a stringent mandatory disclosure regime in their

home country may already provide sufficient voluntary disclosures in their home countrysuch that they do not feel compelled to provide additional voluntary disclosures in the U.S.Given that this is perhaps the first empirical evidence on the Einhorn (2005) model, furtherresearch is recommended in this area on other countries and in other contexts to determinewhether the results are different.

Regarding the second issue, the results failed to reject the hypothesis related to indi-viduality and rejected the third hypothesis (related to power distance) for only one categoryof voluntary disclosures. Therefore, it could be argued that domestic culture affected thevoluntary disclosures of USLAC in the U.S. to only a small extent. This suggests thatthere is some evidence that USLAC are contributing towards international accountingconvergence.

Regarding the third issue, the results of this study failed to reject both the hypothesesrelated to firm size and proportion of foreign sales. The findings indicate that the choiceto use ‘‘international’’ standards was not related to either proportion of foreign revenue orsize of the firm. There are at least two plausible reasons for this finding. First, all thesample companies in this study were listed on a foreign exchange (NYSE or NASDAQ).According to Tarca (2004), the likelihood of a company to voluntarily use ‘‘international’’standards is higher when that company is listed on a foreign stock exchange. Second, asignificantly larger number of companies in this study used ‘‘international’’ standards ascompared to Tarca (2004).23 Therefore, to predict the relation between the characteristicsof those firms and their choice to use ‘‘international’’ standards is difficult. However, when

the definition of ‘‘international’’ standards was split into U.S. GAAP and IFRS, the resultssuggest that companies using U.S. GAAP are larger in size than the companies usingdomestic GAAP. In addition, this analysis shows that companies using U.S. GAAP have alarger proportion of foreign sales than the companies using IFRS. This study has docu-mented that a significant number of USLAC are voluntarily using ‘‘international’’ standards.Therefore, these results, together with the results of culture effects, suggest that the USLACare contributing towards international accounting convergence.

As with all research, there are some limitations of the current study. First, the samplesize of 168 over two years is unevenly distributed among the nine countries. Second, theBotosan (1997) disclosure index, which was developed for U.S. companies, may understatethe level of voluntary disclosures for companies which have lower levels of required dis-

closures in their home countries than in the U.S. Third, the culture scores from Hofstede(2001) were developed before 1980. It is likely that the cultural characteristics of Asiancountries may have changed since then.

Another limitation is that the process for determining a primary independent variable( Disclosure Strictness) has measurement error and it is possible that this affected the results.For example, the conceptualization of the term ‘‘strictness’’ is highly complex. It couldmean ‘‘tightness’’ (i.e., few choices) or it could refer to the magnitude of the penalties forviolating the written rules. In other words, it could be about either the nature of the rulesthemselves, or about their enforcement. We sought to mitigate this limitation in severalways. First, the instrument sent to experts attempted to provide clear instructions regarding

23 Approximately 80 percent companies in this study as compared to 35 percent in Tarca (2004).


24/27


the ranking of the home country mandatory reporting environment (see Appendix) and theusage of the term ‘‘strictness.’’ Second, we provided analysis on the consensus of expertrankings (and it appeared that a reasonable level of consensus was reached). Third, we usedan alternative measure of Disclosure Strictness (the Adhikari and Tondkar [1992] approach)

which was discussed in the Sensitivity Analysis section. This alternative measure yieldedthe same results as the expert measure.An important contribution of the current study is the determination of strictness of

mandatory disclosure regime (disclosure strictness) of the Asian countries. The resultsof Kendall’s W statistic indicated that there was a consensus among the rankings providedby experts. This is perhaps the first study with an empirical finding of this nature and theserankings can be used in the future by researchers in the international accounting area.Additionally, this finding could prove useful for the regulators of the countries which wereranked lower on their mandatory disclosure regime by the experts.

Although the results of the study remained inconclusive about the culture variables, thedesign of voluntary disclosures, strictness of mandatory disclosure regime, and international

accounting convergence are issues that are growing in importance and attracting increasingattention by standard setters and policy makers across the world. Moreover, USLAC are anunder-explored area in the international accounting research. These companies have theirorigin in those Asian countries, which stand at the forefront of economic growth at present(for example, India and China are two of the fastest growing economies and Japan is thesecond largest economy after the U.S.). Therefore, it is important for researchers to obtainempirical evidence on the disclosure practices of these companies. Future research in thisarea may compare the voluntary disclosures by USLAC with matched home country com-panies. Another research question that can be examined is the effect of cross listing onUSLAC. Lastly, the hypotheses tested in this study can be examined in the context of other

countries.


25/27


APPENDIXMANDATORY DISCLOSURE RANKING INSTRUMENT

Please rank the following nine Asian countries from one to nine according to the‘‘strictness’’ of their home country mandatory reporting rules with 1 being the most strict

and 9 being the least strict. Here, mandatory reporting refers to the periodic financial reportsthat the companies are required to submit (such as the annual report), along with therequired content of the report. Some countries have very stringent rules for preparing thesereports and allow the least amount of flexibility in preparing a report whereas other coun-tries allow more flexibility in designing these reports.

Countries (in alphabetical order):China, Hong Kong, India, Indonesia, Japan, Philippines, Singapore, South Korea, Taiwan

Ranking on the basis

of strictness of

mandatory disclosureregime

Country

(most strict)

1.

2.

3.

4.

5.

6.

7.

8.9.

(least strict)

Thank you very much for your time.

REFERENCESAdhikari, A., and R. H. Tondkar. 1992. Environmental factors influencing accounting disclosure re-

quirements of global stock exchanges. Journal of International Financial Management and Accounting 4 (2): 75–105.

Botosan, C. 1997. Disclosure level and the cost of equity capital. The Accounting Review 72 (3): 323–

339.Cahan, S. F, A. Rahman, and H. Perera. 2005. Global diversification and corporate disclosure. Journal

of International Accounting Research 4 (1): 73–93.Choi, F. D. S., C. A. Frost, and G. K. Meek. 2002. International Accounting. Upper Saddle River,

NJ: Prentice Hall.Einhorn, E. 2005. The nature of the interaction between mandatory and voluntary disclosures. Journal

of Accounting Research 43 (4): 593–621.Frankel, R., M. Johnson, and D. J. Skinner. 1999. An empirical examination of conference calls as a

voluntary disclosure medium. Journal of Accounting Research 37 (1): 133–150.Frost, C. A., and G. Pownall. 1994. Accounting disclosure practices in the United States and the

United Kingdom. Journal of Accounting Research 32 (1): 75–102.

———, and W. R. Kinney, Jr. 1996. Disclosure choices of foreign registrants in the United States. Journal of Accounting Research 34 (1): 67–84.


26/27


———, and M. H. Lang. 1996. Foreign companies and U.S. securities markets: Financial reportingpolicy issues and suggestions for research. Accounting Horizons 10 (1): 95–109.

Gray, S. J. 1988. Towards a theory of cultural influence on the development of accounting systemsinternationally. Abacus 24 (1): 1–15.

Haw, I. G., K. Park, D. Qi, and W. Wu. 2003. Audit qualification and timing of earnings announce-

ments: Evidence from China. Auditing: A Journal of Practice & Theory 22 (2): 121–146.Hofstede, G. 1980. Culture’s consequences: International differences in work-related values. Beverly

Hills, CA: Sage Publications.———. 2001. Culture’s consequences: Comparing values, behaviors, institutions, and organizations

across nations. 2nd edition. Thousand Oaks, CA: Sage Publications.Hope, O-K. 2003. Disclosure practices, enforcement of accounting standards, and analysts’ forecast

accuracy: An international study. Journal of Accounting Research 41 (2): 235–272.Hossain, M., H. Perera, and A. Rahman. 1995. Voluntary disclosure in annual reports of New Zealand

Companies. Journal of International Financial Management and Accounting 6 (1): 69–87.Khanna, T., K. G. Palepu, and S. Srinivasan. 2004. Disclosure practices of foreign companies inter-

acting with U.S. markets. Journal of Accounting Research 42 (2): 475–508.

Krantz, M. 2007. U.S. Investors develop crush on U.S.-listed Chinese stocks. USA Today (Septem-ber): 4.

Lang, M., and R. Lundholm. 1993. Cross-sectional determinants of analyst ratings of corporate dis-closures. Journal of Accounting Research 31 (2): 246–271.

Meek, G., and S. Saudagaran. 1990. Survey of research on financial reporting in a transnationalcontext. Journal of Accounting Literature 9: 145–182.

———, C. Roberts, and S. Gray. 1995. Factors influencing voluntary annual report disclosures byU.S., U.K., and Continental European multinational corporations. Journal of International Busi-ness Studies 26 (3): 555–572.

Pett, M. A. 1997. Nonparametric statistics for health care research. Thousand Oaks, CA: SagePublications.

Radebaugh, L., and S. Gray. 1997. International Accounting and Multinational Enterprise. 4th edition.New York, NY: John Wiley & Sons.

Siegel, J. 2005. Can foreign firms bond themselves effectively by renting U.S. securities laws? Journalof Financial Economics 75 (2): 319–359.

Skinner, D. 1994. Why firms voluntarily disclose bad news. Journal of Accounting Research 32 (1):28–60.

Tan, P., S. Goh, L. Tan, and Y. Ting. 2004. Determinants of Internet reporting by Chinese listedcompanies. Working paper, Nanyang Technological University, Singapore.

Tarca, A. 2004. International convergence of accounting practices: Choosing between IAS and U.S.GAAP. Journal of International Financial Management and Accounting 15 (1): 60–91.

Warner, M. 2003. Culture and Management in Asia. London, U.K.: RoutledgeCurzon.Zarzeski, M. T. 1996. Spontaneous harmonization effects of culture and market forces on accounting

disclosure practices. Accounting Horizons 10 (1): 18–37.


27/27

artikel fgd

Documents