Photo by Lora Ohanessian on Unsplash https://bit.ly/3lToo49

Despite a rapidly evolving literature on corporate tax avoidance, relatively little is known about how firms respond to an increase in transparency of their tax affairs. Recent studies examine the impact of mandatory disclosure of tax related information on United Kingdom firms and Australian firms although research on why firms voluntarily disclose tax information and the nature of those disclosures is scarce.

My study addresses this issue by examining how Australian firms respond to a new voluntary disclosure initiative.

The Voluntary Tax Transparency Code

I explore voluntary disclosures made by firms operating in Australia pursuant to the Voluntary Tax Transparency Code developed by the Board of Taxation during 2015-2016 at the request of the Government. The Code seeks to ‘nudge’ subjects towards desired outcomes compared to more ‘command-and-control’ style mandatory disclosure regulations. That is, it encourages but does not compel, firms to publish an annual ‘Tax Transparency Report’ containing certain tax related information not previously available in the public domain, by providing a set of principles and minimum standards to guide disclosure.

The stated aim of the Voluntary Tax Transparency Code is threefold. First, it seeks to have firms publicly disclose more information concerning their tax arrangements to highlight those ‘paying their fair share’. Second, it encourages disclosing firms not to engage in aggressive tax avoidance. Third, it hopes to educate the public about large businesses’ compliance with Australia’s tax laws. The aims were developed during a time of heightened awareness of tax minimisation emanating from the Senate Economics References Committee Inquiry into Corporate Tax Avoidance conducted during 2014-2018.

Tax Transparency Reports provide information in two parts (Part A and Part B). Part A disclosures are primarily numerical items such as a reconciliation of accounting profit to income tax expense/paid/payable, identification of material temporary and non-temporary differences, and effective tax rates. Descriptive disclosures in Part B include the firm’s approach to tax strategy, governance and risk, information pertaining to any international related party dealings, along with a tax contribution summary for taxes paid. Medium businesses (turnover between A$100 million and $500 million) are recommended to disclose Part A, while large businesses (turnover exceeding A$500 million) are encouraged to disclose both Parts.

Hypotheses

Increased transparency of firms’ tax arrangements may be incrementally informative to stakeholders and help firms reduce or eliminate public accusations of aggressive tax avoidance. However, firms may also incur disclosure costs such as reputational damage, proprietary costs, and increased tax authority attention. Thus, firms face a delicate cost-benefit trade-off when deciding to voluntarily provide new information. In theory, firms will provide these disclosures voluntarily when they consider them privately net beneficial.

Pursuant to ‘signalling theory’, some firms may see the Voluntary Tax Transparency Code as an opportunity to confirm their ‘good tax citizen’ status and strengthen their accountability to society at large. Relatedly, ‘legitimacy theory’ posits that a firm’s ability to operate is conditional on their values being perceived to align with those of the society within which it operates. Strategies for firms to achieve legitimacy include information disclosure and acting in a socially desirable way.

To date, empirical research on the determinants and consequences of voluntary disclosure is relatively scant, mainly because few settings exist in which to investigate these. However, it is reasonable to expect larger firms to be more likely to disclose due to them being subject to more regulation and greater public scrutiny which forces them to act in a socially responsible manner consistent with the ‘political cost hypothesis’. This leads to the first hypothesis: the probability of voluntary disclosure of a Tax Transparency Report is increasing in firm size.

Whether voluntary disclosure induces a behavioural change in tax avoidance is unclear and will depend on several factors such as the role of extant tax strategies adopted by the firm or societal attitudes towards tax avoidance and managers’ perceptions of those attitudes. Thus, it is ultimately an empirical question and leads to the second hypothesis: Voluntary Tax Transparency Report disclosers do not change their level of tax avoidance in the years after the introduction of the Voluntary Tax Transparency Code.

Research design and methodology

The sample for the tests of the determinants of voluntary disclosure comprises 263 firms (1,640 observations) while the sample for the tests of changes in tax avoidance comprise 235 firms (1,466 observations). The sample period is 2012-2018 representing 3 years either side of 2015 when the Voluntary Tax Transparency Code was announced.

An attractive feature of this setting is that disclosure of a Tax Transparency Report is voluntary permitting a comparative analysis to be undertaken between disclosers and non-disclosers. To identify the treated firms (namely firms who have published at least one report), a multi-step process was undertaken that relied on two independent sources of information: (i) the Board Of Taxation’s ‘Register of Signatories’; and (ii) the Australian Taxation Office’s Excel file of Tax Transparency Reports.

First, the determinants of voluntary disclosure are examined by estimating a Probit model that regresses a proxy for voluntary disclosure on a set of potential determinants including size, profitability, leverage, capital intensity, intangible assets, market-to-book ratio, the extent to which the firm operates overseas, a dummy variable if the firm has a pre-tax loss, and industry fixed effects. The dependent variable is one of two voluntary disclosure proxies, VTTR_1 or VTTR>1, designed to reflect the extent of voluntary disclosure. VTTR_1 (VTTR>1) is an indicator variable that takes the value of 1 if a firm published one (more than one) report during the sample period, and 0 otherwise.

Second, a difference-in-difference model is estimated using ordinary least squares to examine whether voluntary disclosers change their level of tax avoidance after the Voluntary Tax Transparency Code is implemented relative to a control group of non-disclosers. The dependent variable (TAX) is one of three tax avoidance proxies: the annual cash effective tax rate, the traditional book effective tax rate, and the current book effective tax rate.

Empirical findings

Univariate tests reveal that, on average, voluntary disclosers have significantly higher mean and median cash and book effective tax rates across the entire sample period implying that disclosers are using the disclosure choice as a signal to external stakeholders of their status as ‘good tax citizens’ paying a level of taxes that would likely withstand external scrutiny, consistent with signalling theory and legitimacy theory.

Next, the Probit regression results show that firm size is a strong determinant of disclosure choice consistent with political cost theory. Interestingly, no evidence is found that voluntary disclosure is related to reputational concerns.

The results of the difference-in-differences tests suggest that, on average, voluntary disclosers do not change their level of tax avoidance relative to non-disclosers once voluntary disclosure commences. This is consistent with results of the first research question which suggest that disclosers likely have little, if any, need to reduce undesirable tax avoidance practices.

Interestingly though, evidence is found of a decrease in tax avoidance by non-disclosers suggesting the Voluntary Tax Transparency Code is successful in encouraging some firms to engage in less tax avoidance.

Conclusion and policy implications

The effectiveness of tax transparency as a corrective tool has been questioned recently. So, investigating the Voluntary Tax Transparency Code’s recent introduction advances our knowledge and understanding of the relatively unexplored area of how firms respond to calls for voluntary disclosure of their tax affairs.

Overall, the results suggest the Code merely acts as a confirmatory mechanism for large firms already paying relatively higher taxes consistent with several theories on voluntary disclosure. However, the decision to voluntarily disclose is not accompanied by a reduction in tax avoidance by disclosing firms relative to non-disclosers consistent with arguments that greater mandatory or voluntary disclosure is not a panacea for addressing corporate tax avoidance.

For instance, Holland et al. (p. 338) conclude ‘Perhaps if society wishes to change corporate behaviour it should do so via changes in legislation aimed at reducing the opportunity and benefits of tax avoidance rather than attempting to use increased disclosure as a means of changing managers’ behaviour.’

Nevertheless, evidence is found that non-disclosing firms decrease their level of tax avoidance perhaps to avoid public scrutiny, or in anticipation of the Voluntary Tax Transparency Code becoming mandatory soon. Thus, the Code may have the unintended effect of shaping corporate behaviour in an indirect way highlighting a possible latent benefit of transparency as an accountability mechanism.

The findings are timely given the ongoing post-implementation review of the Voluntary Tax Transparency Code (announced in February 2019) and may inform global policymakers currently deliberating the introduction of voluntary disclosure regimes to supplement existing mandatory disclosure requirements.

This article has 2 comments

  1. I am interested in one of the conclusions « The results of the difference-in-differences tests suggest that, on average, voluntary disclosers do not change their level of tax avoidance relative to non-disclosers once voluntary disclosure commences. » What is the measure for the level of tax avoidance? For example, if a corporate is paying 30% tax or more that no change in that effective rate means that there is no change in their level of tax avoidance?

    • Dear Andrew, thanks for your comment. The paper uses 3 measures for the level of tax avoidance: (1) the annual cash ETR (CETR) defined as cash taxes paid per the cash flow statement divided by pre-tax income from the income statement; (2) the traditional book ETR (BETR) defined as total income tax expense divided by pre-tax income; and (3) the current book ETR (CBETR) defined as current income tax expense divided by pre-tax income. Your interpretation is correct – lower (higher) ETRs imply higher (lower) levels of tax avoidance.

Leave a comment

Your email address will not be published. Required fields are marked *

*