فرهنگ سازمانی شرکت های حسابداری عمومی : شواهد از شرکت های محلی تایوانی و شرکت های وابسته ایالات متحده

This study examines the organizational cultures of public accounting firms with data from US affiliated international accounting firms in Taiwan and Taiwanese local firms. Hypotheses are tested about the impact of the national culture of the US firms on their Taiwanese affiliates, and about cultural differences across function and rank. The study extends previous research by (1) examining an oriental country whose culture is significantly different from that of the US, (2) using the Hofstede, G., Neuijen, B., Ohayv, D. D., & Sanders G. (1990). Measuring organizational cultures: a qualitative and quantitative study across twenty cases. Administrative Science Quarterly, 35, 286–316 practices-based measure of organizational culture, and (3) examining the importance of the fit between employee preferences and organizational culture in influencing organizational commitment, job satisfaction and propensity to remain with the organization. Support is found for the cultural impact and fit hypotheses. Additonally, while culture is found to be relatively homogeneous across function, differences are found across rank.

The organizational culture of public accounting firms is an important area of research because of culture's ability to affect the motivation, behaviour and performance of accounting firm participants generally (see, for example, Holmes & Marsden, 1996, Hood & Koberg, 1991 and Pratt & Beaulieu, 1992), and specific decisions and actions including auditor judgment on materiality (Carpenter, Dirsmith, & Gupta, 1994) and auditor independence (Windsor & Ashkanasy, 1996). In an era of increasing globalization, attention has also been given to the impact of organizational culture on accounting firm mergers across national boundaries (Pratt et al., 1993 and Radebaugh & Gray, 1997), and the affiliation (or association) of international accounting firms with local firms in different nations (Iqbal et al., 1997 and Soeters & Schreuder, 1988). Noting that the “big” international accounting firms are essentially US in orientation, Soeters and Schreuder (1988) studied the extent to which US national culture influenced the organizational culture of US firms operating in the Netherlands with Dutch employees. Similarly, Pratt et al. (1993) examined the extent to which the US culture of international accounting firms was inculcated in the value systems of British and Australian employees of those firms in Britain and Australia. Soeters and Schreuder (1988, p. 83) and Pratt et al. (1993, p. 627) pointed to the tensions and difficulties that may arise in merger and affiliation activities resulting from the problem of “culture clash”. The importance of this issue remains, and indeed has increased, since these studies. With respect to mergers, the “Big Eight” of Soeters and Schreuder in 1988 had become the “Big Six” of Pratt et al. by 1993 and, by 1998, the “Big Five”. With respect to affiliations, national ownership and/or other legislative restrictions mean that for some nations (e.g. Taiwan, Indonesia and Japan), affiliation or association with a local accounting firm is the only way that an international firm can establish operations in that country (Radebaugh & Gray, 1997, p. 651). From another perspective, the current Vision Project of the AICPA, in recognizing the broader and higher levels of service provisioning of CPAs, also emphasises the importance of globalization, identifying international services as one of the top five core services of the CPA profession and citing global forces as a key issue in their “Profession-wide Visioning Process” (AICPA, 2001, pp. 07–08htm). The literature also raises the issue of subcultures in accounting firms. Both Hood and Koberg (1991) and Pratt and Beaulieu (1992) suggest that culture may differ across function (specifically, management advisory services (MAS) and non-MAS functions of audit and tax) and rank. Cultural differences between audit and MAS functions were observed as long ago as Watson (1975), while Carpenter et al. (1994) also found evidence of differences across rank. The question of whether organizational subcultures exist is important. If culture differs across function, this may cause problems for organizational design, coordination and control. Pratt and Beaulieu (1992, p. 681) suggest that these problems are possible reasons why some firms have chosen to “spin off” their MAS function. If culture varies across rank, then, as Hofstede (1998, p. 11) points out, there is the potential for decisions made at a higher level, and based on that level's perception of organizational reality, to produce conflicts and adverse consequences when those decisions impact lower levels with different perceptions of reality. In examining the potential influence of the US culture of the international accounting firms on their affiliates in Taiwan, the current study extends prior research in three ways. First, Pratt et al. (1993), who studied the influence of US culture on British and Australian employees of US international accounting firms in Britain and Australia, noted that both these countries were, like the USA, Anglo in cultural heritage with English as their national language. They were also countries where “the US presence is strong and large US firms are well entrenched” (Pratt et al., 1993, p. 627). Hence, culture transference might be more readily and easily achieved (and, hence, less visible as constituting transference) in these country combinations than where the combinations involve greater cultural differences. Pratt et al. (1993, p. 628) noted this as a limiting factor in their study, and called for future research to be conducted in oriental countries because the greater differences between US and oriental cultures may make cultural influences more difficult to achieve (and, hence, more visible as cultural influences). Taiwanese culture is very different from that of the US and the cultures of Britain, Australia and the Netherlands, the countries at issue in the Pratt et al. (1993) and Soeters and Schreuder (1988) studies. For example, taking two of the most important cultural characteristics differentiating western and oriental cultures, individualism and power distance (Lachman, Nedd, & Hinings, 1994, p. 49), Hofstede (1983, p. 82) places the USA, Britain, Australia and the Netherlands in the small power distance and high individualism quadrant cluster of countries, and Taiwan in the opposite quadrant of large power distance and low individualism. Hence, as compared to prior studies, the current study can provide stronger tests of the effect of US affiliation on the organizational culture of accounting firms located outside the USA. Second, this study extends previous research by using the Hofstede, Neuijen, Ohayv, and Sanders (1990) practices-based measure of organizational culture. Prior studies have typically used values-based measures of culture, in particular Hofstede's (1980) and Hofstede and Bond's (1988) measure comprising the five dimensions of power distance, individualism, uncertainty avoidance, masculinity and Confucian dynamism. However, these measures were originally developed to measure culture at the national level, leaving open the question of “whether it is appropriate for use within and among organizations” (Pratt & Beaulieu, 1992, p. 670). Pratt et al. (1993, p. 627) stated that future research “could usefully benefit from using the measure of organizational culture recently developed by Hofstede et al. (1990)”. The third way this study advances prior research is that it tests for effects of the fit between employees' preferences and the firm's organizational culture on outcome variables of organizational commitment, job satisfaction, propensity to remain with the firm, and performance. These latter variables have been consistently identified in the general management literature and in studies of public accounting firms as being important individually and organizationally valued outcomes (Ketchand & Strawser, 1998 and O'Reilly et al., 1991). Holmes and Marsden (1996, p. 29) observe that the absence of tests for such effects is a major omission in prior research on accounting firm culture.

With respect to the question of whether the organizational cultures of the “Big Five”, essentially US, international accounting firms may influence the organizational culture of affiliated firms in different countries, our results support those of Soeters and Schreuder (1988) and Pratt et al. (1993), which suggest that such influence is present. Specifically, our results show differences in the organizational cultures of the Taiwanese local firms and the Taiwanese affiliates of the “Big Five” firms, with the local firms demonstrating organizational culture characteristics consistent with Taiwanese national cultural attributes, but the affiliates demonstrating organizational culture characteristics more consistent with US national culture.6 Hence, these findings provide evidence of the ability of the international firms to transplant their organizational practices, which are consistent with and supported by their (US-based) culture, into their overseas affiliates. As noted at the outset, the previous studies in this area, particularly Pratt et al. (1993), involved US affiliates in countries with cultures relatively similar to those of the USA and, hence, where cultural influence and transference might be more easily achieved. An important contribution of this study is that it shows that such cultural influence and transference is evident even where the national culture of the affiliate (here, the oriental culture of Taiwan) is substantially different from that of the USA. Thus, our study adds evidence not only of the presence of cultural influence of the “Big Five”, US based, international firms found in previous studies, but also of the strength of such influence across significant national cultural divides. With respect to function, consistent with Hood and Koberg (1991) we found culture to be relatively homogenous across function. This might reflect a commonality in the general backgrounds and training of accounting firm employees prior to functional specialization, or the over-arching nature of public accounting firm responsibilities and societal expectations. If supported by future research, our results suggest that this homogeneity may not require different organizational structures for different functions, or different coordination and control processes, at least insofar as structures and processes are affected by cultural considerations. This is not meant to simplify, or to suggest closure on, the issues of accounting firm structure or control processes either across or within functions. Gupta, Umanath, and Dirsmith (1999), for example, discuss the complexity and differentiation within the audit function structure associated with audit task characteristics and audit social processes including team dependence relations and size. With respect to rank, our findings indicate that, compared to lower levels, higher level managers perceive organizational cultures as being more employee (vs job) oriented, and more parochial (vs professional). These findings are, again, broadly consistent with Hood and Koberg (1991) and Pratt and Beaulieu (1992) in the public accounting firm context, and Hofstede (1998) in a commercial firm context. Taken together, the findings suggest that this difference in perception between top and lower levels may be a general phenomenon, and one to which top managers need to be sensitive. If top managers believe that their organization's culture and practices are people oriented and involving, but lower-level employees perceive otherwise, there is potential for misunderstanding on both sides and for adverse consequences for behaviour. Another important finding of the study was the association between the fit of an individual's preferred and perceived actual culture with the outcome variables of organizational commitment, job satisfaction and propensity to remain with the organization. Our results suggest that commitment, satisfaction and propensity to remain are enhanced when the organizational culture of the firm is congruent with the employee's preferred culture. These findings have implications for accounting firms, domestically and internationally, in their employee selection (and socialization) practices, and in their design of management and organizational practices which give rise to, and mould, their organizations' climate and culture. The findings of our study are also important for cross-cultural research in accounting. Recent reviews of that research (e.g. Chow et al., 1999 and Harrison & McKinnon, 1999) have stressed the need to go beyond the aggregate cross-national level of analysis and to explore cultural differences at the firm and sub-unit levels as well as the national level. The findings of our study provide further impetus to this call in showing the potential insights from examining the nature and impact of culture at a range of different levels. Some limitations of our study need to be recognized. First, we have relied on prior studies to support the supposition that the US accounting firms' organizational culture would be affected by US national culture. We do not have direct measures of the US firms' domestic organizational culture to corroborate this. This was also the assumption relied on by others, including Soeters and Schreuder (1988), for example, and is defensible particularly in view that our findings are consistent with this assumption. Nonetheless, direct exploration of this issue would usefully be part of future research. Second, we cannot rule out the possibility that the Taiwanese US affiliates may have self-selected for affiliation. To investigate this, however, would require longitudinal or before/after comparisons of organizational culture which, in turn, would require knowledge of which firms would subsequently become affiliated at some future time. These requirements would make such investigation scarcely feasible. A final limitation is, as noted earlier, that the Cronbach alpha reliability coefficients for the individual practices (Ps) comprising Hofstede et al.'s (1990) organizational culture dimensions were low. This is largely a function of the small number of items for each dimension because, while alpha is a frequently used reliability measure, it is affected by the number of items in the scale, favouring scales with greater rather than fewer items. Confidence in the Hofstede et al. (1990) measure was otherwise provided by acceptable performance against the item to scale total correlation diagnostic, and by the fact that our results for the affiliation, function and rank hypotheses were broadly consistent both with those hypotheses and with prior studies which have used different measures of organizational culture.