سایه های خاکستری : بررسی تجربی مدل گری فرهنگ و شیوه های اندازه گیری درآمد با استفاده از اطلاعات 20 - f

Gray (1988) developed a theory linking accounting values and systems to Hofstede’s (1980) cultural dimensions. To date, no studies have used actual reported data to test the Gray model. This study addresses this shortcoming by using data from the SEC Form 20-F to test proposed relationships between Gray's (1988) accounting value of conservatism (in income measurement practices) and Hofstede's (1980) cultural values. It also tests three other non-cultural environmental variables (tax rates, relative size of capital markets, and influence of the European Union) proposed by Gray (1988) and Salter and Niswander (1995) as modifiers to cultural effects. The results of this study suggest that, as theorized by Gray (1988), the cultural variable individualism is significantly and positively related to differences in income measurement practices between countries. This study also finds that a country's membership in the EU and corporate tax rate is related to income measurement practice differences. However, unlike previous studies, this study does not find that the source of equity is related to income measurement practices.

Gray's (1988) model of the influence of social values on accounting is one of the seminal articles in international accounting. A review of the citation index and list for Gray (1988) in Google Scholar indicates some 502 cites of this study with its nearest competitor having less than 200.2 One dimension of Gray's (1988) model, the accounting value of conservatism, is particularly important and yet remains relatively untested (Doupnik and Tsakumis, 2004). As Gray (1988:10) observes, conservatism would seem to be a significant accounting value because “it is arguably the most ancient and probably the most pervasive principle of accounting valuation.” Doupnik and Tsakumis (2004) find it the most important area of Gray (1988) remaining to be explained. Doupnik and Tsakumis (2004:35) suggest “that researchers concentrate their efforts on conservatism. Conservatism is the accounting value that affects the recognition and measurement of items that appear in financial statements and therefore would have the greatest implication for the cross-national comparability of financial statements.” This study proposes to take up the challenge laid down in Doupnik and Tsakumis (2004) and search for the causes of unconditional conservatism. Drawing on the work of Hofstede, 1980, Nobes, 1987, Gray, 1988, Salter & Niswander, 1995 and Doupnik & Tsakumis, 2004 we propose (1) that a country's degree of conservatism is a reflection of national societal values and norms, and (2) in determining the level of conservatism, the effects of societal values are supplemented by its institutional structure

Using data from approximately 3089 Form 20-F reconciliations from 1998–2004, this study finds that: 1. On average, companies from more individualistic countries report less conservative income numbers. This is in line with Gray's (1988) propositions. Surprisingly, the study also finds that countries with higher power distance also report higher income numbers. This fills the first gap in the knowledge. 2. In filling the second gap (the effect of institutions): a. As expected countries with relatively high tax rates report more conservative income numbers. b. Contrary to prior studies (Adhikari & Tondkar, 1992 and Salter & Niswander, 1995), the underlying importance of equity as a finance source in firms' home markets no longer bears a significant relationship to reported income. This is in line with the Ball et al. (2008) finding that unconditional conservatism is not related to the importance of national equity markets. 3. After controlling for cultural effects, countries that are members of the EU report income numbers that are consistently less conservative income numbers than non-EU countries. This is contrary to the previous high levels of conservatism reported in EU countries at the beginning of our test period where continental EU members were conservative and the UK and Ireland optimistic (negatively conservative). It also indicates the ability of extra national institutional factors to affect income measurement outcomes (conservatism) thereby filling the third knowledge gap. This study adds value not only in its results, but also in a number of methodological and data improvements on previous work. They are: 1. Using actual financial reporting data to measure conservatism. Previous studies have used proxies for conservatism such as auditor perceptions of practices (Salter and Niswander, 1995) or non-accounting financial data (Sudarwan and Fogarty, 1996). Building on the findings of Ball et al. (2003) that there are often significant differences between prescribed GAAP and actual reporting practices, actual financial statement data is used to derive the dependant variable. 2. Using audited data. The quality of this data is enhanced by the SEC requirement that an independent Certified Public Accountant audit the data. 3. Using ratios. Raw differences in income are converted into Gray's (1980) indices of conservatism thereby removing biases of firm size. 4. Using data from countries representing a majority of capital markets. The countries in the sample represent 74%, by value, of the total market capitalization of the world's stock markets in 2004. Starting with Gray's (1988) model (as modified by Doupnik and Tsakumis, 2004), this study links societal culture and institutions with accounting values and practices. It then tests the ongoing validity of the model across four cultural constructs, five institutional variables and seven years. We find that while individualism continues to be a powerful force in predicting preferences for optimistic reporting, other factors such as tax rates and membership in the EU can push countries without optimistic cultural predilections to move to optimism in income measurement. The results show that, in the end, Gray's (1988) model, as amended by Doupnik and Tsakumis (2004), continues to have good predictive ability but that the institutional portions of that model have increasing value over its base cultural dimensions. Perhaps this paper's most significant result is that the EU, an institution with little impact on accounting at the time of Gray's (1988) model, is now a very powerful force for convergence. The question remains however, now that IFRS is in force and is practiced in the EU, whether the EU will remain a force for global convergence or become a force for regional convergence and global divergence. To that end this paper represents a historical benchmark for future tests of IFRS influence in Europe. Every study has limitations and perhaps the biggest limitation to this study is the self-selection bias of the firms. The only firms that are included in this study are those that chose to list on U.S. stock markets and file a Form 20-F reconciliation. Typically these are large firms that want access to U.S. capital and they may not be representative of the smaller firms from their country. Larson and Street (2004), for example, provide evidence that in the European Union there is a small and large GAAP being practiced. However, the argument can be made that this self-selection bias actually decreases the likelihood of finding any results. The foreign firms that list on U.S stock exchanges may be more likely to have smaller reconciliation values due to choosing income measurement practices that are closer to U.S. GAAP. Thus, any findings of cultural relativity may have added power. A second limitation involves the use of national tax rates. It is possible that the national tax rates used in this study do not reflect the effective tax rate of firms. For example, a corporation may be taxed at the national level and then again at a local level, and perhaps more. Finally, this data is the result of reporting on the cusp of mandatory IFRS use in Europe. We have already noted that this knocked Germany out of our sample. While we could have updated the data to collect such data it is our feeling that the world is in an accounting transition and, as pointed out by Linthicum (2008) and others, no one quite knows what the impact of IFRS on accounting outcomes will be. This study provides potential for future research in related areas. It focuses on reconciliation of the net income numbers of foreign firms that list on U.S. stock exchanges. Those firms that provide an income reconciliation are also required to provide a reconciliation of the equity section from foreign GAAP to U.S. GAAP. To the extent that equity is affected by entries that pass directly to the equity section without impacting income, a study that examines differences in reconciling equity from foreign GAAP to U.S. GAAP could provide further evidence on which cultural values are related to financial reporting. Second, as data was collected it appeared that many firms chose to report in US GAAP rather than reconcile. The motivation for such behavior would seem an interesting extension to this study. Third, this database is, in itself, a hand collected example of the income measurement of some 3000+ firm years over a seven year period. Perhaps the greatest opportunity for future research may come from moving from the national or macro level to the individual firm or micro level. This would permit the inclusion of determinants such as industry and firm size in examining the measurement model. It is entirely possible, for example, that in certain industries there are no cross-national differences and in others no similarities. This study would, of itself, have to grapple with the issue of how culture could be operationalized at the firm level although it is possible that the legal system could be a substitute. Finally, as mentioned in the limitations, this data is the result of reporting on the cusp of mandatory IFRS use in Europe. What now becomes interesting is if this transition means a single global GAAP relatively similar or different from the US or simply a series of culturally shaded IFRS GAAPS.