Tax Complexity and Firm Tax Evasion: A Cross-Country Investigation

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Review

Manuscript ID economies-2952117

“Tax Complexity and Firm Tax Evasion: A Cross-Country Investigation”

Economies

The impact of tax complexity on tax evasion remains an unresolved empirical issue, despite the often expressed and widely held notion that complexity leads to more tax evasion. This paper uses World Bank data that captures some elements both of tax evasion and of tax complexity to estimate the determinants tax evasion. The authors find that their measures of complexity are positively correlated with their measures of tax evasion, a result that holds up under a variety of robustness tests.

These are plausible results, with the obvious implication that “Addressing the complexity of the tax system could potentially mitigate the adverse impact on tax evasion levels in these countries.”

The paper is clearly written, it is based on a useful data set and a sound methodology, and its results are (reasonably) plausible, with many robustness tests.  

Let me raise several issues that I believe should be addressed.

First, the obvious and main problem is the absence of any clear identification strategy, so that the estimation results demonstrate correlation, not causation. Now the authors acknowledge this problem themselves, but not until the very end of the paper. I do not see an easy way to fix this problem, and its presence could easily be a justification for rejection.  I am inclined to be a bit more forgiving here, given the other contributions of the paper. Even so, the authors need to make this limitation much clearer and much earlier in the paper.

Second, the data set seems a bit odd. As far as I can tell from Table 2, the number of high income countries (especially OECD) countries is quite limited; for example – and this is not a complete list – countries like Denmark, France, Italy, Austria, Australia, Japan, US, UK, the Netherlands, … are not included, for reasons that are not obvious.  These (and many other countries) are certainly included in the Doing Business data, and I assume that they are also included in the Enterprise Survey data, although I could be wrong on both counts.  In any event, the author should clarify the reasons for the specific selection of countries.

Third, it is worth mentioning that the World Bank stopped the Doing Business survey for reasons related to (potential) favorable treatment of some countries, notably China. I do not think that this negates the value of the entire data set, but it should be mentioned.

Fourth, the basic specification of any “determinants of tax evasion” equation should include the basic features of the fiscal system that the taxpayers face. In particular, measures of enforcement (audit rates, fine rates) should be included.  These variables are not included here.  I doubt that such variables can be compiled/constructed for the full sample of countries. However, there are plausible proxies for the enforcement regime, and in fact the authors already include such proxies in the form of such control variables as “Government Effectiveness”, “Regulatory Quality”, “Rule of Law”, and “Control of Corruption”. I suggest that the authors reframe these control variables as proxies (even if imperfect proxies) for the enforcement regime. In this regard, I would also have thought that the authors would include the widely available measure of “tax morale” as an additional control variable. There are clearly potential endogeneity issues with many of these variables, but their inclusion would be consistent with other studies of the determinants.

Fifth, there are some puzzling results. As the authors write regarding the main results in Table 5:

“Surprisingly, voice and accountability, and the rule of law consistently 594 emerge with significant unexpected signs. In addition, the unemployment rate and gender 595 have mixed coefficient signs.

However, the authors make no attempt to explain these results.  It is not enough merely to note the presence of unexpected results. Some attempt should be made to explain these results.

Sixth, I wonder whether there are different effects of complexity on tax evasion by sector, as suggested by Table 3.  I like the many robustness tests of the authors, and I suggest that the authors also explore further the potential impact of complexity on tax evasion by sector.

Finally, on a purely and minor comment, I think that the authors should display a bit more modesty – the phrase “boldly ventured” seems to me an overstatement of the paper’s contribution.

So, I appreciate and value the obvious efforts that the authors put into the paper. I believe that the paper can be strengthened by addressing the above comments. 

Author Response

Dear Reviewer,Thanks for the thorough review of our manuscript, the encouraging words, and the helpful comments. As indicated in the file, we have addressed all your valuable comments. All the paragraphs we quote in the file (written in red) refer to the revised manuscript. We hope that our revisions are enough to convince you to recommend our article for publication.

Regards,

Authors

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This is a really excellent paper. The one comment I had was whether to clarify the relationship not just to illegal tax evasion but also to legal tax avoidance. For example, the very complex OECD transfer pricing guidelines are the basis for tax avoidance in developing countries.

Author Response

Dear Reviewer,

Thanks for the thorough review of our manuscript, the encouraging words, and the helpful comments. As indicated in the file, we have addressed all your valuable comments. We hope that our responses are enough to convince you to recommend our article for publication.

Best Regards,

Authors 

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

In my opinion, it is a very high-quality text that is suitable for publication.

I have only a few minor comments:

„Each country has approximately 65 to 24,000 firms as representatives from 27 industries.“ It could be reformulated.

„Countries with inconsistent rankings for two measures of tax avoidance (the absolute value of the discrepancy between the two ranks exceeds 70)“ Should be explained - you compare 0/1 variable vs %.

 „Third, we find a significant positive correlation (p-value < 1%) between the shadow economy measures compiled by Schneider et al. (2010) dan??? Elgin and Öztunali 287 (2014) and the tax evasion ratio.“ Check the spelling.

„Referring to the literature review section, 𝛼1 is expected to be negative and significant, indicating that a more complex tax system is associated with a lower incidence of firm tax evasion and a lower firm tax evasion ratio.“ Check the meaning (positive!). 

Note - There is/was an opinion at the Ministry of Finance of the Czech Republic that the data for the Czech Republic in the Doing Business database is inaccurate, misleading, and suspicious, with reference to an interesting fact - the Czech Republic and Slovakia had (have) very similar systems (initially the Czechoslovak tax system from 1993), but very different tax values time (see table 2, 866 vs 325).

Author Response

Dear Reviewer,

Thanks for the thorough review of our manuscript, the encouraging words, and the helpful comments. As indicated in the file, we have addressed all your valuable comments. All the paragraphs we quote in the file (written in red) refer to the revised manuscript. We hope that our revisions are enough to convince you to recommend our article for publication.

Regards,

Authros

Author Response File: Author Response.pdf