【正文】 Section 2. I then make some observations about their particular paper in Section 3. Conclusions are reached in . As well as adding some caveats to the ?ndings of Ding et al., this paper might be helpful to future users of the data in GAAP 2021. 2. The data . Fit for purpose? Ding et al. (2021, p. 3) refer to the use of Price Waterhouse (PW) data in prior research, which includes that by da Costa et al. (1978), Frank (1979), and Nair and Frank (1980). Nobes (1981) had earlier noted that it is dangerous to use these data for academic research because, among other problems, they were not designed for the purpose. Does use of the data in GAAP 2021 suffer from this problem? Although it is not reported in GAAP 2021, the motivation for that survey was to protect large accounting ?rms from criticism (by the World Bank and others) resulting from the then recent collapse of panies and economies in the Far East. The survey ai med to reveal the existence of the large differences from IAS (or absences of requirements pared to IAS) in the accounting rules of many countries so that poor reporting would not be blamed on poor auditing. The objective was to focus the attention of regulators in any particular country on improving accounting rules rather than on attacking the audit profession. As such, the survey?s purpose was not to enable international parisons, let alone to provide data for academic research. Nevertheless, as long as there are no systematic biases in the data, it might be reasonable to use them for research. For example, whereas the PW data started from a questionnaire that focused on differences between US and UK accounting (thus highlighting differences between these two countries), I am not aware of any such national bias in GAAP 2021. The reference point for parisons was International Accounting Standards (IAS), which is a bias, but this need not affect the purpose of Ding et al. This bias is discussed later (see Section ). . Rules not practices In addition to the national bias in the PWdata, a further problem noted in Nobes (1981) is that differences in the rules (de jure differences) are mixed with those relating to practices (de facto differences). How does the GAAP 2021 data pare? GAAP 2021 does not suffer from this problem. It records only de jure differences between national and IAS rules, not de facto differences between national and IAS practice. Although not so serious a limitation as would be created by mixing rules and practices, the concentration in GAAP 2021 on rules rather than practices could cause problems for research, which Ding et al. do not discuss. For example,if a nation?s rules do not require a particular item to be disclosed but panies often disclose it in practice, then this ??absence” of a rule should perhaps be ignored. Or, if a national system (unlike IAS 38) allows internallygenerated research costs to be capitalized but in practice panies do not capitalize, then the ??divergence” in rules is perhaps irrelevant. Another aspect of this is that some de jure differences do not lead to de facto differences in a particular country because the issue is irrelevant. For example, the absence of rules on pension accounting is of little importance in China because Chinese panies do not generally run de?ned bene?t pension plans. More subtly, both ??inconsistency” categories in GAAP 2021 (see the ?rst paragraph of this paper)contain two types of inconsistency with IAS: (i) where the national rule and the IAS is inpatible (. if the national rule required LIFO but IAS required FIFO), and (ii) where the national rule would not ensure IAS pliance (. if the national rule allowed either LIFO or FIFO, but IAS required FIFO).The former inconsistency is more serious. Indeed, the latter may be of no practical importance (. if panies using the national rule choose not to use LIFO). . An IAS bias The GAAP 2021 data were based on looking at accounting rules from one direction: the content of IAS. So, if a national system had more rules or more restrictive rules than IAS had, this did not show up. For example, US GAAP covered many issues on which IAS was silent (. oil and gas accounting)。 and UK GAAP did not allow LIFO whereas IAS did. Since these types of difference are not covered by GAAP 2021, they were not included by Ding et al. (as they note in their Appendix A). If these differences were included, it would make the US and the UK look more different from IAS than the ??absence” and ??divergence” measures suggest, but it would not much affect the position of the Netherlands, where the rules were generally less detailed or less restrictive than IAS. . 111 topics but 79 survey questions Ding et al. (in Appendix A)