The relationship between tax transparency and tax avoidance

Abstract

All over the world, the tax avoidance practices of large multinational firms have received much media attention over the last few years, which has become a prominent reputational risk for many firms. In addition to the possible reputational risk stemming from corporate tax avoidance, these tax practices can also have dire consequences for the economies in which these firms operate. Global tax transparency initiatives were developed in an attempt to address the issues created by global tax avoidance. There is, however, little academic evidence on whether increased tax transparency can have an effect on corporate tax avoidance. The purpose of this study is therefore to investigate the relationship between tax transparency and tax avoidance. A content analysis was firstly used to qualitatively assess the extent of tax transparency disclosures in the annual corporate reports of the top 100 firms listed on the JSE. Thereafter, a regression analysis was used to determine the relationship between tax transparency and tax avoidance. Tax transparency scores were used as a proxy to measure tax transparency while both effective tax rates and cash effective tax rates were used as a proxy to measure tax avoidance. The study finds that firms which are more transparent in the disclosure of their tax affairs also have higher effective tax rates and cash effective tax rates.

Keywords:

• cash effective tax rate
• effective tax rate
• tax transparency
• tax avoidance
• reputational risk
• voluntary disclosure

Acknowledgements

This work was financially supported by PWC. The content analysis of the tax transparency reporting was based on the PWC tax transparency framework that was made available to the authors.

Notes

1 The G20 is a forum of the world’s major economies that is made up of 19 countries and the European Union. The 19 countries are Argentina, Australia, Brazil, Canada, China, France, Germany, India, Indonesia, Italy, Japan, Mexico, Russia, Saudi Arabia, South Africa, South Korea, Turkey, the United Kingdom and the United States. (G20, n.d.).

2 This database is available on a subscription basis.

3 The Pearson correlations are based on the raw value of the observations and evaluates the linear relationship between two continuous variables and is linear when a change in one variable is associated with a proportional change in the other variable (Burns & Burns, 2008). The Spearman correlations are based on the ranked values of the observations and evaluates the monotonic relationship between two variables that tend to change together, but not necessarily at a constant rate (Burns & Burns, 2008).

4 A perfect model fit refers to the ‘line of best fit’ that minimizes the sum of the squared deviations from the line to the dots on the scatter graph, thus errors of prediction will be at a minimum (Burns & Burns, 2008).

5 The coefficient of determination, known as the R-square in a multiple regression, enables one to assess how well the regression equation fits the actual data. The R-square is interpreted as how much of the total variation in the dependent variable (Y) is explained by the variance of the independent variables (X). The higher the R-square, the better the line fits the data as more variance is explained (Burns & Burns, 2008). R-square is the square of the r-correlation as shown in Table 3 that measures the strength of a linear relationship between two variables (x and y). The r-correlation can take a value of +1.00 to −1.00, while a value of zero indicates that there is no association between the two variables. A perfect positive correlation is indicated by +1.00 and a perfect negative correlation is indicated by −1.00 (Burns & Burns, 2008).

6 The t-statistic is used to decide whether the null hypothesis should be accepted or rejected. If the t-statistic reaches the conventional significance levels, the null hypotheses will be rejected which means that it is unlikely that data observations happen by chance (Burns & Burns, 2008).