The perceived fairness of turnover tax

Estimates of the contribution of small, medium and microenterprises (SMMEs) to the economy vary. In terms of contribution to gross domestic product (GDP), an estimate of 52% – 57% has been quoted by the Deputy Minister of Trade and Industry, Elizabeth Thabethe, who put the number of SMMEs in South Africa at 2.8 million and their contribution to employment at 60%. Moreover, looking ahead, the National Development Plan projects that, by 2030, no less than 90% of new jobs will be created in small and expanding firms (Groepe 2015).


Introduction
Estimates of the contribution of small, medium and microenterprises (SMMEs) to the economy vary.In terms of contribution to gross domestic product (GDP), an estimate of 52% -57% has been quoted by the Deputy Minister of Trade and Industry, Elizabeth Thabethe, who put the number of SMMEs in South Africa at 2.8 million and their contribution to employment at 60%.Moreover, looking ahead, the National Development Plan projects that, by 2030, no less than 90% of new jobs will be created in small and expanding firms (Groepe 2015).
The South African Revenue Service (SARS) (2008) reported that many small businesses do not pay tax as they are overwhelmed by the complexities of the current tax system and high compliance cost, and this was confirmed in later research (Smulders et al. 2016).
In order to address the problem of the complexities of the current tax system and high compliance cost, South Africa introduced a simplified tax system for businesses, individuals or companies with a turnover which does not exceed an amount of R1 million (SARS 2008).In terms of the provisions of par 2 of the Sixth Schedule to the Income Tax Act No. 58 of 1962 (the Act) the R1 million is for the entire year of assessment.The simplified tax system, called Turnover Tax, was implemented with effect from 01 March 2009, applicable in respect of years of assessment commencing on or after that date (s 48A of the Act).Turnover Tax, for qualifying taxpayers, substituted normal Income Tax, Provisional Tax and Secondary Tax on Companies (STC) (STC became Dividend Tax with effect from 01 April 2012) (SARS 2011).Since South Africa implemented the simplified Turnover Tax system for microbusinesses in 2009, only about 7 700 businesses have registered for this tax and of these only 12% are new taxpayers: This means only about 924 new businesses registered as taxpayers under the Turnover Tax system (National Treasury 2011a:75).The Davis Tax Committee (2014:26) reported that as at 04 July 2013 there were only 7827 active microbusinesses and Visser (2016) reported that it is still less than 10 000.
Is it possible that the reason the number of taxpayers who have joined the Turnover Tax system has not been more prolific is that the Turnover Tax system is not fair?This article reports on research conducted to explore the perceived fairness of the current Turnover Tax system from the perspectives of both the government and small business owners.In order to address this problem better, the characteristics of a fair tax system, as advanced by Smith (1776), were examined against the criticisms and provisions of the Turnover Tax system by way of a correspondence survey.The research sought to fill a gap by investigating whether the current method of taxing microbusinesses by way of the Turnover Tax system in line with the provisions of the Sixth Schedule to the Act is perceived to be a fair method to tax a business.It was advantageous to reaffirm whether the The South Africa Turnover Tax system, implemented on 01 March 2009 to simplify tax for microbusinesses and to improve tax compliance had an insignificant number of registrations and research indicated that a possible reason is the fairness of the tax.The aim of this study is to explore the perceived fairness of the Turnover Tax system.By way of a literature review, criticisms and provisions of the Sixth Schedule to the Income Tax Act No. 58 of 1962 were identified and used as statements on a survey questionnaire.Using the principles of a fair tax system as advanced by Adam Smith a correspondence survey which included two open-ended questions was issued to participants with knowledge of Turnover Tax to establish whether the statements corresponded to any of Adam Smith's Maxims.The results reveal that the Turnover Tax system is not perceived as completely fair and encourage Government to relook at the legislation with the intend to simplify it further, to removing ambiguity and add detailed lists of excluded services.Education and training of taxpayers is important.A repetition of this study on taxpayers registered on the Turnover Tax system will make further contribution and add to the insight into the fairness of the Turnover Tax system provided by this article.principle of fairness, as discussed by Adam Smith (1776), is enhanced or undermined with regard to Turnover Tax from the points of view of the businesses (Hayek 1960).In order to add to our understanding of the perceived fairness of Turnover Tax, the study interpreted the results of a correspondence survey conducted using participants who understand tax and assist businesses with tax matters.The analysis of the results of the correspondence survey, including the remarks on the two open-ended questions attached to the correspondence survey, was used to further our insight into whether specific arguments and counterarguments, identified by way of a literature review, are within Smith's (1776) definition of tax fairness.

Methodology
This study adopted an inductive approach to explore each of the elements of the Turnover Tax system which may be regarded as unfair, as discussed in the literature review.A survey questionnaire was developed on the basis of the literature review in order to shed light on the perceived fairness of Turnover Tax (Ahrens & Chapman 2006;Maroun, Turner & Sartorius 2011).Prior to the survey being distributed to participants, the survey was piloted.A purposeful selection technique was applied with the view of selecting only registered and knowledgeable tax accountants from auditing, accounting and law firms, as well as tax academics, based in South Africa, in this way helping to gather a diverse range of respondents involved with advising businesses on various tax matters and who have, based on their qualifications, a knowledge of both the maxims of Smith (1776) as well as Turnover Tax (Creswell 2009;Houghton, Jubb & Kend 2011).A comparable approach was used by Brivot and Gendron (2011) and Cohen, Krishnamoorthy and Wright (2002), who carried out similar interpretative studies aimed at exploring perceptions.To ensure reliability, a sample of 60 tax experts was selected (Bendixen 1996;Maroun et al. 2011).Purposeful selection using a small sample size enhances the quality of the research findings by ensuring that all participants have knowledge of Turnover Tax (Creswell 2009;Maroun et al. 2011).
The research method was a correspondence analysis.The correspondence survey included open-ended questions at the end, allowing the researchers to gather information in a way which might not have been possible if only structured statements were used.The open-ended questions often reaffirmed what was being asked in the structured statements and sometimes provided a different response because they allowed participants to express their opinions without being restricted by the statement and were able to make possible recommendations.Such recommendations could be considered by SARS for possible review or implementation in order to make the Turnover Tax more attractive or fairer to taxpayer's (Bazeley 2002;Creswell 2009;Kalof, Dan & Dietz 2008).
Participation in the survey was voluntary.The researcher contacted the potential participants via telephone or e-mail, explained the purpose of the study to them and asked whether they wished to participate in the survey.If they were willing, the researcher sent them a copy of the correspondence survey for completion, with guidelines for how to complete it.Should any statement, in their opinion, not be related to any of Smith's maxims, the respondent was requested to leave it blank.The open-ended questions needed to be read and answered accordingly.A copy of the survey questionnaire is presented in Appendix 1. Once all surveys were completed, the data were collected and aggregated into a contingency table.When aggregating the data, each 'X' was allocated a value of 1 and where the block was blank, no value was allocated.The data from the correspondence analysis were then analysed using the STATA software programme with the aid of a statistics expert.Results were generated and a two-dimensional graph plot was created.Each of the maxims was plotted on a separate axis and the number from each corresponding statement was then plotted in a space on the graph depending on each statement's correlation coefficient and inertial contribution to the respective axes.The correspondence analysis was, as a result, an easy way to interpret correlation between individual statements and elements of fairness (Smith 1776).
The inertia (variance) attributed to each cell in the contingency table was calculated using the row profiles and masses.Using principal component analysis, the dimensions (axes) as well as the point-rows and point-columns were defined (Bendixen 1996;Maroun et al. 2011).The inertial contribution of the fairness criteria needed to be considered in order to determine whether the specific criteria should be considered in the analysis.An average inertial contribution of 25% (100% divided by four criteria) was calculated.Fairness criteria one and four were above the 25% limit and so were included (Appendix 2).Fairness criteria two and three were 19.6% and 19.9% (Appendix 2), respectively, and so would have been excluded from the discussion; however, fairness criteria two and three combined made up the meaningful portion, 39.5%, of the total inertia and were therefore still used in the analysis because the combination of the contribution was still significant.The respective axes were determined based on the correlation coefficients of the fairness criteria, the inertia of the fairness criteria and the coordinates of the criteria.The respective axes summarised in Table 1 were based on the above-mentioned bi-plot interpretation.
Similarly, the inertial contribution of the row categories needed to be ascertained in order to establish whether the category formed part of the analysis.Any statement with inertia greater than the inertial contribution of 4.5% (100% divided by 22 statements) was considered as part of the analysis and the rest of the statements were not analysed further in the study (Bendixen 1996;Maroun et al. 2011).The statements shaded in Appendix 2 indicate which of the statements formed part of the study analysis after considering the inertial contribution.The coordinates, the correlation coefficients and the inertia of the statements were considered in the analysis and interpretation of the results.The correlation coefficients were then used to determine whether the statement was correlated to dimension two (x-axis) or dimension one (y-axis).It was important to note that the sign of the correlation had no indication of the strength of the relationship but only acted as an indicator of which dimension on the axes the statement relates to (Bendixen 1996;Maroun et al. 2011).The inertia was also considered to decide whether the statement contributed a significant amount to the dimension.In the case where a statement strongly correlated, however trivial the inertia might have been, a conclusion needed to be drawn, stating the fact and indicating that the data would not provide the most meaningful result.

Structured open-ended questions
The responses to the open-ended questions were captured, analysed and used to complement and add to the credibility and richness of the data obtained from the correspondence analysis.If there were any differing opinions, the differences were noted and discussed.Possible recommendations were also offered by respondents on how to improve the perceived fairness of the Turnover Tax system.

Fairness of turnover tax
Perceptions of fairness are instrumental to evaluate the sense of justice, equality and the credibility of a tax system (Bird & Zolt 2003;Hartner et al. 2008;Maroun et al. 2011).Smith's (1776) four maxims of a fair tax provide a guide to what a tax system should incorporate in order to ensure its perceived fairness.In this literature review, Smith's (1776) four maxims were used as a basis to determine whether the Turnover Tax system, which had been in existence for the last six years, complies with the fairness principles.

An overview of the meaning of 'fairness' using Smith's maxim
Adam Smith is considered by some to be the father of modern political economy.In 1776 The Wealth of the Nations was published, in which Smith argued that there were four maxims of a fair tax system which, when combined, would create a fair and effective tax system, and those same maxims still form the basis for a fair tax system (  2009).For the purposes of the research, it was deemed appropriate to use Smith's (1776) maxims as a guide to assess the fairness of the Turnover Tax system.The unfairness aspects identified in the literature review were translated into a survey questionnaire which was used as a correspondence tool in order to assess people's perceptions on the fairness of the Turnover Tax system (Appendix 1).

Smith's maxim 1: Equity
According to Smith's (1776) first maxim (maxim 1), tax should be paid in proportion to any taxpayer's ability to pay the tax to the government.Wealthier individuals should be required to pay more than poorer individuals (Akrani 2010).In order to determine whether the Turnover Tax system complies with the fairness principles in relation to maxim 1, the authors considered two main aspects: the taxpayers should be able to afford the tax due and the tax should be charged based on revenue (Maroun et al. 2011;Vivian 2006).
The first consideration which requires that taxpayers should be able to afford the tax is not promoted because a taxpayer whose income is generated by way of a certain profession listed under the professional services list would be prohibited from registering for Turnover Tax (para 3 of the Sixth Schedule) even if the qualifying turnover is less than R1 million (BDO Spencer Steward 2009;International Finance Corporation 2007;Willemse 2010).Further disqualifications include, but are not limited to, capital asset disposals in excess of R1.5 million over a 3-year period, individuals who trade by way of a partnership and will only be allowed to register if the partnership's turnover does not exceed R1 million, as well as taxpayers whose total income consists of more than 20% of investment income or if the taxpayer trades by way of a personal service provider or labour broker (S5, S6 and S7, Appendix 1) (BDO Spencer Steward 2009).
Another area of concern is that, in terms of the provisions of s 48B of the Act, all sectors are taxed at the same rate, despite different sectors having different expenses and profit margins, as well as start-up businesses which hardly make any profit because of costs; and so, taxing all sectors at the same rate may not be fair (S2, S15, S17, S18 and S19, Appendix 1) (BDO Spencer Steward 2009;Hassan 2009).
The second consideration of maxim 1 indicates that tax should be levied based on revenue.Revenue is defined in terms of the Oxford Dictionary of Accounting as 'any form of income' (Oxford 2010).The fact that dividends are excluded creates a contradiction to maxim 1, as a dividend constitutes income (S12).Taxable turnover also excludes non-business income, for example, remuneration and investment income, even though the revenue definition would include this (S20) (Hassan 2009).However, rental income is included as part of taxable turnover, which is appropriate (S11) (Viviers 2009) (para 6 and para 7 of the Sixth Schedule to the Act).
Arguments in favour of maxim 1 are that Turnover Tax is an elective tax and so the taxpayer is given a choice and one can argue that this takes into consideration the taxpayers ability to bear-and not the government's attempt to collect-revenue (S8 Appendix 1) (BDO Spencer Steward 2009; SARS 2010).Government argues that the Turnover Tax rates are very low and so should compensate for the fact that tax is payable on turnover and not profit, and also that it promotes the ability to bear principle (S3 and S16, Appendix 1) (International Finance Corporation 2007).

Smith's Maxim 2: Certainty
Smith's (1776) second maxim (maxim 2) requires that the tax which is being levied should be certain and there should be no elements of subjectivity or confusion.Prior to the tax being paid, the taxpayer needs to be certain about how much should be paid, when it should be paid and how it should be paid, not only for the tax professional but also for the person who is being taxed (Akrani 2010;Economic Concepts 2010;Lambert 1992).The Turnover Tax system was implemented with the aim of simplifying tax for taxpayers and reducing the number of different taxes payable by the same taxpayer, promoting maxim 2 in that it makes tax easier to comply with (S4 Appendix 1).By reducing the number of taxes that individuals and business need to comply with and simplifying the rules within Turnover Tax, it has become more certain, yet expecting individuals to pay different taxes on nonbusiness income (remuneration and investment income) (S20 Appendix 1) does not promote maxim 2 (Hassan 2009).
The fact that Turnover Tax requires tax to be paid based on a cash basis (S6) contradicts maxim 1, but it simplifies the payment of tax and so promotes maxim 2. The fact that there are not many complex exemptions and deductions, that the tax is payable at the same rate and that all businesses of a single taxpayer are added together to pay one tax, increases certainty and simplicity (S17, S18 and S19) (Hassan 2009;International Finance Corporation 2007).
The professional services as defined in par 1 of the Sixth Schedule to the Act and excluded from qualifying as a microbusiness per the provisions of par (3)(b)(i) and (ii) of the Sixth Schedule to the Act lists broad areas of professional services.Willemse (2010) argues that there is ambiguity involved because the list only contains the broad areas and the allocation of specific services to the broader service is open to interpretation and does not promote maxim 2 (S1, S9 and S10, Appendix 1).Engel (2016) also expresses his concerns with the fairness of the exclusion of professional services because he says that all microbusinesses face the same problems and challenges, irrespective of whether their services are regarded as professional services, and should be treated the same.
Microbusinesses and individuals who wish to register for Turnover Tax are at times not sufficiently educated about whether Turnover Tax is a feasible solution or whether they even qualify for Turnover Tax (S13 Appendix 1) (Hassan 2009).Government has made a pledge to increase training and education (National Treasury 2011b) and it is also recommended by the Davis Tax Committee (2014) and once implemented would promote maxim 2.

Smith's Maxim 3: Convenience
Smith's (1776) third maxim (maxim 3) requires convenience; in other words, the method and timing of paying tax should be convenient for the taxpayer.If the tax is convenient and easy to pay, then individuals and businesses will be able to pay the tax and tax evasion will be reduced (Akrani 2010).Turnover Tax is required to be paid on a biannual basis, in terms of the provisions of para 11 of the Sixth Schedule to the Act, and because it is mainly individuals, micro type businesses or start-up companies which do not have large sums of money, biannual payments may add up to a substantial amount and businesses may not have sufficient funds to make these payments.It may, therefore, not be convenient for taxpayers to pay these taxes every 6 months and maxim 3 is then not promoted (S14 Appendix 1) (Hassan 2009).The change to para 9 of the Sixth Schedule to the Act, which allows a microbusiness to deregister voluntarily from Turnover Tax (S22), and the change to the provision of the Sixth Schedule to the Act to allow a microbusiness to remain registered for VAT while it meets the requirements to pay tax on a turnover basis for income tax (S21 Appendix 1) promote maxim 3.

Smith's Maxim 4: Simplicity or efficiency
Smith's (1776) fourth maxim (maxim 4) requires that tax administration surrounding the payment of tax should be simple and easy to comply with and that expenses relating to the collection of taxes, for example, compliance costs, should be affordable and as low as possible for the taxpayer (Akrani 2010;Economic Concepts 2010).The fact that Turnover Tax reduces the number of taxes and, as a result, reduces the tax returns needing to be completed, that all sectors within the Turnover Tax system are taxed at the same tax rate, that the system does not allow for deductions against the taxable turnover and that only a few exemptions are allowed and that the tax is on turnover will promote maxim 4 (S2, S3, S4, S16 and S18, Appendix 1) (BDO Spencer Steward 2009;Hassan 2009;International Finance Corporation 2007;Willemse 2010).However, this may result in a contradiction of maxim 1, as discussed in 3.2.1.On the other hand, the fact that certain parts of the legislation lack detail-for example, the professional services list (para 3 of the Sixth Schedule to the Act)-and that the lack of detail on the professional service list might be open to interpretation does not promote maxim 4 (Willemse 2010).Lack of knowledge by the taxpayer and lack of training provided by government about the Turnover Tax system may also make it difficult for taxpayers to understand, resulting in complicated tax issues, for example, whether a microbusiness qualifies to register resulting in professional help being required and increasing costs (S1, S5, S9 and S13) (Hassan 2009).This will not promote maxim 4.

Results
The correspondence analysis results indicating the relationship between Smith's (1776) four maxims and the Turnover Tax fairness statements are discussed next.The findings of the two structured open-ended questions will only be provided with the discussion of the results in part 5 of the article because the responses to the open-ended questions were closely aligned with the results of the correspondence analysis.

Correspondence analysis
The findings of the correspondence analysis were presented by way of a correspondence bi-plot, illustrated in Figure 1, developed in order to show the relationship between Smith's (1776) maxims and the Turnover Tax fairness statements identified and used in Table 1. Figure 1 plots only those statements with a significant inertia which is greater than 4.5% (discussed in the method section).Each maxim had to be plotted and interpreted on its own axis, creating a fourdimensional figure.This would result in a 100% retention value, meaning that no information was lost in the results: however, as this four-dimensional figure was not possible given the available technology, a two-dimensional figure was created.Figure 1 shows that this two-dimensional bi-plot has a retention value of 79.02%, which indicates that the results are explanatory in a significant way (Bendixen 1996;Maroun et al. 2011).
For the purposes of defining the axes of the correspondence plot in Figure 1, three fairness criteria as labelled in Table 1 are illustrated in this figure.An average inertial contribution of 25% (100% divided by four criteria) was calculated.Fairness criteria one (equity) and four (simplicity or efficiency) were above the 25% limit and were included.Fairness criteria two (certainty) and three (convenience) were 19.6% and 19.9%, respectively, and so should have been excluded from the discussion.However, these two criteria made up a significant portion (39.5% of the total inertia) and so were still used in the analysis as in combination the contribution was still significant and was plotted on a negative X-axis (axis 1) in Figure 1.Any statement with inertia greater than the inertial contribution of 4.5% (100% divided by 22 statements) was considered to be part of the analysis, and all other statements were not analysed further in the research.(Bendixen 1996;Maroun et al. 2011).The statements plotted in Figure 1 indicate which of the statements formed part of the research analysis after considering the inertial contribution.
Statements dealing with the disqualifications of individuals providing professional services, the disqualification provisions relating to shareholding in professional service entities and other investments, as well as the criteria for individual partners in a partnership to register for Turnover Tax (S1, S5 and S6) were closely clustered and were positioned almost on the positive y-axis, indicating a close correlation among the three statements and fairness maxim 1, namely that taxpayers contribute in proportion to their ability.Taxpayers with the same turnover will have the same ability to pay irrespective of the kind of industry he, she or it works in and it is unfair to treat professional services differently.The correlation coefficients of the statements with the y-axis were 0.914, 0.847 and 0.872.There was, however, a low correlation of 0.016, 0.121 and 0.127 to the undefined positive x-axis, demonstrating no meaningful relationship, and this did not need any further discussion.
Conversely, statements dealing with the equal tax rates for high-, low-or no-profit margins, the disallowance of any deductions for expenditure incurred and the exclusion from Turnover Tax of certain sectors (S10, S17 and S18) were clustered close to the negative y-axis, indicating a close correlation between the three statements and fairness maxim 4, namely that the tax paid by taxpayers ought to be as little as possible.The correlation coefficients of the statements with the positive y-axis were 0.564, 0.62 and 0.691, indicating a close correlation between the three statements and fairness maxim 1, namely that the taxpayers should pay tax in proportion to their ability to pay tax.The statements also had a correlation of 0.436, 0.298 and 0.374, respectively, to the undefined positive x-axis but indicated no real correlation to any of the other fairness criteria.
Non-business income (remuneration and investment income), which was excluded from the taxable turnover of a microbusiness and taxed as per the normal tax tables in the hands of the individual (S20), was strongly correlated to the undefined positive x-axis with a correlation coefficient of 0.84 and a relatively high inertial contribution of 14.6%, indicating that the statement was not promoting any of Smith's (1776) fairness criteria.The corresponding correlation coefficient to the negative y-axis was 0.16; however, the contribution was only 1.9%, indicating that a small sample felt this statement related to maxim 4, namely that the tax paid by taxpayers ought to be as little as possible.
Statements which dealt with the inclusion of rental income as part of investment income, the exclusion of dividends from taxable turnover and the requirement for Turnover Tax taxpayers to pay tax biannually (S11, S12 and S14) were strongly correlated to the negative y-axis, indicating a close correlation between the three statements and fairness maxim 2, namely taxpayers' certainty that tax will be payable and maxim 3, namely that tax will be paid at a time or in a manner that is convenient to the taxpayer.The respective correlation coefficients of the statements with the y-axis were 0.289, 0.255 and 0.546.On the negative y-axis, there was a meaningless correlation between the statements and maxim 4, namely that the tax paid by taxpayers ought to be as little as possible (0.024, 0.167 and 0.005, respectively).

Discussion
This section of the article further discusses the results of the correspondence analyses by using Smith's (1776) maxims as the base.The responses are incorporated in the discussion in order to add further insight into the findings of the research.

Maxim 1
The literature review suggested that the list of all factors and professions, included in para 3 of the Sixth Schedule to the Act, which would disqualify a taxpayer from registering for the Turnover Tax system (S1, S6 and S5), might be perceived as unfair because it excludes taxpayers with low income from paying tax on the turnover they receive.The literature further suggested that the list of professional services excluded from using the Turnover Tax system might be open to interpretation and taxpayers will have to get a professional to interpret the provisions.This is not promoting maxim 2, dealing with certainty of paying tax, or maxim 4, dealing with simplicity of the tax system (S1 and S5).The results proved first of all that the respondents felt that the statements mainly related to maxim 1 dealing with the equity of the tax system and, secondly, that the respondents perceived none of the identified and tested statements to be unfair.

Maxims 2 and 3
The literature review suggested that S11, which includes rental income as part of taxable turnover, and S12, which excludes dividends (par 6[b] of the Sixth Schedule to the Act) from taxable turnover, pose a contradiction to equity maxim 1, because both (rental and dividend) sources of income are revenue in nature and should be treated the same.The respondents strongly associated the different treatment of rental versus dividends with the belief that for the tax to be fair and certain (maxim 2) there should be no elements of subjectivity or confusion in the Turnover Tax system, and if the tax treats two sources of income (rental and dividend) differently it contradicts the fairness and in particular the certainty (maxim 2).The results indicated that the respondents did not feel that it was fair for rental income to be classified as part of investment income when looking at the taxable turnover of a taxpayer registered on the Turnover Tax system.Results also indicated that respondents perceived that it was not fair that dividends do not form part of taxable turnover.These findings were in line with unfairness aspects identified in the literature review.
The literature also suggested that the fact that turnover taxpayers are required to pay interim payments on a biannual basis (S14) was not supportive of the belief that the method and timing of paying tax should be convenient (maxim 3), because taxpayers need to pay tax in advance based on turnover generated even though they might not yet have the cash to pay the tax.The results indicated that the fact that Turnover Tax taxpayers were required to pay interim payments on a biannual basis was not promoting convenience (maxim 3) and respondents were not wholly confident of the fact that payments of Turnover Tax on a biannual basis were fair.This was emphasised by the fact that respondent 12, in the open-ended questions, stated that the number of payments should be reduced to one per annum.

Maxim 4
The literature review suggested that the disqualification of professional services to register for Turnover Tax is very limiting as certain sectors are prevented from registering for the tax, even if their qualifying turnover is below the R1 million threshold.This substantially reduces the businesses that can register for Turnover Tax (S10) and is not supportive of the certainty maxim (2) because potential Turnover Tax taxpayers first have to evaluate whether their trade is a professional service before considering whether they can register.The respondents were uncertain whether this statement (S10) related to any of the maxims, and this uncertainty was verified in the responses to the open-ended questions where four respondents specifically commented on the statement.Respondent 17 indicated that there was 'a definite need to add more detail to the professional service definition'.Respondent 35 answered that 'there should be less restrictions on persons permitted to register for Turnover Tax, especially with regard to the disqualification of all professional services'.Respondent 25 wrote that 'taxpayers should not be limited just because they are a professional service firm: they can still as small businesses contribute to economic growth'.Respondent 26 wrote that perhaps SARS should try to simplify the Turnover Tax system even more, implying, among other things, that the reference to professional services, without guidelines to help understand what constitutes it, is perhaps too complex and difficult to interpret by a person starting a business with no understanding of tax or business.
The answers to the question on what SARS can do to improve the Turnover Tax system identified a potential area in which SARS needs to make adjustments and that is to the professional services definition in para 1 of the Sixth Schedule to the Act by making it less limiting, by disqualifying fewer services, being more specific in what is included in each item of professional services, ensuring no ambiguities and perhaps supplying a list of businesses that qualify as a professional service.The list of professional services can be added to the Sixth Schedule to the Act and can be subject to review and change at the discretion of the Minister of Finance.All the answers received from respondents confirmed that the disqualification of all professional services and other exclusions as per para 3 of the Sixth Schedule to the Act does not promote the fairness of the Turnover Tax system.
Should a concern or person qualify as a microbusiness (par 2 of the Sixth Schedule to the Act) such concern or person, irrespective of its profit margin, will pay tax at the same rates (S17) and will not be allowed to reduce its taxable turnover with any deductions for any expenses incurred on start-up of the business or trade (S18).In the literature, these statements were found to be supportive of the certainty maxim (2) but not of the equity maxim (1) because this results in persons being taxed at the same rate, despite having different profit margins and expenditures.The results revealed that respondents perceive the fact that Turnover Tax is at the same rate without any additional deductions to promote the simplicity maxim (4).One respondent was commenting on S17.Respondent 8 stated that different rates should be applied for different-sized businesses or some form of special sliding scale should be implemented.With regards to S18 respondents throughout the survey noted, in the open-ended questions, that there was a need for a certain set of exceptional costs to be allowed as deductions by taxpayers registered on the Turnover Tax system.An example given by respondent 14 was: 'What if you get robbed and some of your cash gets stolen, how will tax be payable?'Currently, the Turnover Tax system would still require the tax to be paid and the question is, is that fair in a country with a high rate of theft?

Statement not associated with any maxim
'Non-business income (remuneration and investment income) is excluded from the taxable turnover of a microbusiness and taxed as per normal in the hands of the individual (S20)' was the only statement that was convincingly not associated with any of the four maxims.

Conclusion
The main aim of the article was to establish whether the Turnover Tax system was perceived to be fair.The authors established a basis for testing perceptions by way of identifying possible 'issues' relating to the Turnover Tax system from prior literature, assessed it against Smith's (1776) maxims and translated the issues into a survey questionnaire used to assess taxpayers' perceptions on the fairness of the Turnover Tax system.In the analysis of the results of the correspondence survey, it was noted that only a small number of the statements had some correlation with Smith's (1776) maxims of fairness; however, there was no perfect correlation found and some statements had no correlation at all.This implies that Turnover Tax is potentially not being perceived as totally fair.
On analysing the recommendations, respondents felt that taxpayers registered for Turnover Tax who make a loss or minimal profit should not be required to pay tax by way of special provisions within the legislation, for example, some sort of deduction.Under the Turnover Tax regime, the accumulated losses are also lost forever where a taxpayer decides to rather close the business, whereas in the normal tax regime the unincorporated individual would have closed the business with personal accumulated tax losses to be utilised going forward (Visser 2016).Surely taxpayers will not elect a system where they may be worse off.
Although suggestions have been made that different tax rates should be considered for different business sectors as different sectors have different profit margins, the key to this kind of system should be simplicity and this was also listed by Smulders (2014) as the primary criteria.With simplicity in mind, the timing of tax payments needs to be reconsidered, either having only one payment to reduce administrative burdens or having more payments more often to reduce cash flow burdens.Respondents also felt that Turnover Tax is not simplified enough and National Treasury needs to consider the redrafting of provisions relating to registration requirements, qualifying thresholds, disqualification criteria and the publication of a more detailed list of professional services in order to be less ambiguous and to allow more potential taxpayers to register for Turnover Tax.However, the most prevalent recommendation from the open-ended questions was to improve the promotion of the Turnover Tax system and to increase the training and education given to potential Turnover Tax taxpayers, as this will ultimately result in improved compliance and a higher number of registrations.This is such an important recommendation because, as Groepe (2015) said, 'to grow the number of small businesses in the country the education system has an important role to play'.Education will not only raise the level of skills among the workforce, it will also play a role in empowering potential entrepreneurs to realise their own potential.Smulders ( 2014) also recommended that government should look at implementing taxation as a school subject, invest money in training and workshops to educate persons starting businesses and to investigate the set-up of a Government Department to help and support micro-, small and medium businesses.
The correspondence analysis has added to the debate with regard to the perceived fairness of Turnover Tax.The intention of the study was never to quantify or generalise the findings but rather to inform perceptions and gain an understanding from multiple perspectives.The inherent limitations of this study mean that a definitive conclusion with reasonable assurance is not possible as we are looking at individual's perceptions.Nevertheless, the results obtained from the study add to the credibility of the issues identified in the literature available and, in some instances, promote the opinions of the literature.
In a recent study, it was confirmed that with the current provisions for Turnover Tax in place small businesses carry a higher proportionate burden in respect of tax and employees' compliance activities (Smulders et al. 2016).This confirms that the Government will have to relook at innovative ways to reduce compliance costs of microbusiness in order to encourage growth that will stimulate employment and assist the Government with South Africa's high unemployment rate.
A repetition of this study on taxpayers registered on the Turnover Tax system will make a valuable further contribution to the insights gained from this study.Future research on simplified tax systems in other developing countries might also provide SARS with valuable insights.
This article contributes to the debate on the fairness of the Turnover Tax system and leaves us with the following thought: Tax is certain, but is it fair?
Small business forms part of the backbone of a thriving society and it should be supported with fair tax practices in order to grow and support it.Dividends do not form part of taxable turnover (Viviers 2009).
13 Microbusinesses are at times insufficiently educated to make qualified decisions about whether to register (Hassan 2009).
14 Turnover Taxpayers are required to pay interim payments on a biannual basis (Hassan 2009 19 Where a person trades in more than one unrelated business, the total turnover of all the business activities is added together in order to determine the qualifying turnover cap (Hassan 2009).

20
Nonbusiness income (remuneration and investment income) is excluded from the taxable turnover of a microbusiness and taxed as per normal in the hands of the individual (Hassan 2009).

21
A proposal in the 2011 Budget Speech will allow a microbusiness to be registered for VAT and Turnover Tax simultaneously (National Treasury 2011b).

22
A proposal in the 2011 Budget Speech will allow a microbusiness to voluntarily deregister from Turnover Tax at any time (International Finance Corporation 2007).
The structured open-ended questions were the following: 1. What do you think South African Revenue Service should implement or change with regard to the Turnover Tax system in order to make the system more attractive to microbusinesses? 2. What do you think South African Revenue Service should implement or change with regard to the Turnover Tax system in order to make the system fairer to microbusinesses?

FIGURE 1 :
FIGURE 1: Bi-plot illustrates the relationship between statements and fairness criteria.
low profit margins are taxed at the same rates as businesses with high profit margins (BDO Spencer Steward 2009).18Nodeductions are awarded for any expenses incurred (BDO Spencer Steward 2009; International Finance Corporation 2007).