90756 Investment Climate l World Bank Group Nuts & Bolts Technical Guidance for Reform Implementation Can Specialized Tax Software Reduce Tax Compliance Costs in Developing Countries? World Bank Group The World Bank Group conducted a study, using data from a number may spend substantial amounts of time in experimenting with different of surveys of businesses in developing countries, to investigate the scenarios for the sake of minimizing association between the use of specialized tax software and total the tax bill. Nevertheless, these are tax compliance costs. Most empirical data from both developed and general assumptions, and they warrant more rigorous scrutiny to gain a better developing countries show that use of tax software is associated understanding of the details of tax with higher tax compliance costs. However, since larger and more compliance costs in developing countries. sophisticated businesses are likely to both have higher tax compliance costs and use tax software, it is necessary to control for these and Specialized tax software packages are designed to help firms with other firm characteristics. Regression analysis shows that it is indeed computing their tax bills. They are often possible for tax software to reduce tax compliance costs, holding complementary to bookkeeping or other variables constant, but the results are weak and often mixed. accounting software packages. General accounting software packages (mainly for Many taxpayers apparently “over-use” the software in attempting to bookkeeping, such as Quicken™ in the optimize their tax liabilities. United States) are used to keep track of business activities including daily sales, inventories, payroll, and so on. Firms then At first glance, it might be assumed that without software” and 40.1 hours per transfer all the final balances (profits, the availability and use of specialized return for “software preparation.”1 sales, and costs) to tax accounting tax software for tax compliance would software which could be specialized tend to reduce taxpayers’ compliance Some of the difference may result from (such as Turbo Tax™) or non-specialized costs (in addition to other benefits the simple matter of selection bias: An (such as Excel™) to compute the final tax such as reduction in errors), compared earlier article (Coolidge 2012) which bill payable to the government. These with the alternative of making manual documents findings from developing software packages may be free (or easily calculations involving relatively large countries reveals that relatively larger pirated) in some countries or come at amounts of data. However, available and more complex businesses are likely a cost in others (for instance, about empirical evidence seems to show that to have higher tax compliance costs $150 for TurboTax™ for small business taxpayers using such software spend due to the sheer volume of relevant federal income tax in the United States2). more time, on average, than those transactions, and they are also more Nevertheless, the cost of learning and not using it. For example, DeLuca et likely to be computerized, have skilled adopting the software may be more al. (2007) provide data showing that labor, and therefore invest in specialized important than the cost of purchasing self-preparers in the United States with tax software. The difference may also the software (appropriately amortized) in no software spend an average of 161 exist as a result of the relative ease of affecting tax compliance costs. to 164 hours per year in tax compliance using the software to help with more while those with software spend an “tax planning” (for example, running Previously, it was assumed that using average of 234 to 287 hours per year. several simulations of different scenarios specialized tax software packages may On a “per return” basis, Guyton et al. based on different categorizations of decrease tax compliance costs because (2003) estimated an average of 18.2 expenses or timing of realization of they make the computation process hours per return for “self-preparation losses). It is not unlikely that tax staff fairly easy for firms, helping them avoid Can Specialized Tax Software Reduce Tax Compliance Costs in Developing Countries? 1 PERCENTAGE OF ARMENIAN FIRMS USING TAX FIGURE 1:  with tax software usage likely to appear as a benefit for SOFTWARE BY TURNOVER (IN ARMENIAN DRAMS) firms? The first step was to divide the main research question into two parts: (i) Which firms are most likely to adopt tax more than 58.35 million software?, and (ii) How much do tax compliance costs differ between tax software users and non-users? 30 to 58.35 million What types of firms are most likely to use 5 to 29.99 million specialized tax software? The results of the analysis outline several patterns in tax up to 5 million software usage and firms’ characteristics (Yılmaz and Coolidge 2013). The firms more likely to adopt tax software are located 0 20 40 60 80 100 in relatively more developed (urbanized) regions of a country, Percent larger in size (measured by turnover or total employment), established under more complex legal forms, paying more types Users Non-users of taxes, operating in relatively more capital-intensive industries, Source: IFC 2011 [Armenia]. and doing most of their tax compliance in-house. Sample data Note: At the 2009 exchange rate, one U.S. dollar was equal to 363.28 Armenian drams. from Armenia3 are illustrative, as shown in Figures 1 and 2. Survey findings on firms’ perceptions of the business PERCENTAGE OF ARMENIAN FIRMS USING TAX FIGURE 2:  environment shed light on how important perception SOFTWARE BY NUMBER (OF TAXES PAID) indicators are in shaping firms’ decisions with respect to tax software usage in tax compliance. The most important 3 indicators include: 4 • how costly it is to keep records for tax purposes; 5 • how hard is it to find workers with proper skills and education; 6 • actual education and occupation (as a proxy for productivity) 7 and more of the “most knowledgeable person” about tax compliance in a firm (that is, who was asked to respond to the questions 0 20 40 60 80 100 in the survey); and Percent • formality in business activities (approximated by the percentage of turnover transacted via bank accounts). Users Non-users Source: IFC 2011 [Armenia]. Results show that firms that do not find it “difficult” to keep records for tax purposes are more likely to use tax software than those who report finding it difficult. This may alternatively computational errors which could be very costly, especially imply that firms who use tax software face less difficulty in taking into account the higher likelihood (and costliness) keeping records for tax purposes. As firms report less difficulty of audits that might result from such errors (which are not in finding skilled workers, they become more likely to use tax included in this analysis). Overall, the costs and benefits of software to help them comply. In most cases, the education and using tax software depend on the purchase price of the occupation of the tax staff (the most knowledgeable person in software, how it is adopted, and how accounting staff are tax accounting in each firm) are highly and positively correlated educated and become proficient in its usage (for instance, with the usage of tax software, as shown in Figure 3. Finally, the degree of tax planning or productivity of the accountants) firms that are more formal in their business activities (that is, over time, including short-run versus long-run effects. those that transact a higher percentage of their turnover via Consequently, it is not clear whether one should expect a bank accounts) are also more frequent users of tax software. negative or positive association—if an association exists and is large enough to capture statistically—between tax software Do tax compliance costs differ between usage and tax compliance costs. specialized software users and non-users? The study conducted by the World Bank Group focused on the For each country, the study compared total tax compliance costs effect of using specialized tax software on tax compliance costs of software users with that of non-users (in terms of money for small and medium enterprises in developing countries. In and time separately) and then looked at different variables that particular, is a reduction in tax compliance costs associated influence both the decision to use tax software and the level 2 Investment Climate l World Bank Group of tax compliance costs. The results from this simple analysis PERCENTAGE OF ARMENIAN FIRMS USING TAX FIGURE 3:  show that specialized software users generally pay higher tax SOFTWARE (BY EDUCATION) compliance costs. The difference is quite large and economically important. However, this difference cannot be attributed to tax Ph.D software usage alone, without controlling for all important firm Graduate characteristics. In a comparison of the tax compliance costs paid degree by tax software users and non-users in the same turnover group, Undergraduate the difference was much smaller. In one case (Armenia), the degree study found that tax software users on average face lower tax compliance costs than non-users, for firms in the same turnover Specialized secondary education band. However, the difference was not statistically significant. Secondary school Thus, determining whether tax compliance costs differ between software users and non-users is not simple. Because the decision to adopt tax software is not exogenous, any analysis 0 20 40 60 80 100 must control for the main firm characteristics (to overcome Percent selection problems) before inferences about the association between tax software and compliance costs can be made. Users Non-users Source: IFC 2011 [Armenia]. Note: Education is based on that of the most knowledgeable person in the firm regarding tax Do firms using specialized tax software have compliance. lower tax compliance costs than similar firms that do not use it? Other control variables generally had predictable results in the This analysis held various firm characteristics constant. In analysis. Firm size, paying value-added or profit tax, and the the case of Nepal, and focusing on the firms that do all tax number of types of taxes paid were each associated with a compliance fully in-house (that is, not outsourcing any work significant increase in tax compliance costs. This finding was to external tax preparers), tax software usage is not associated statistically significant and robust. Moreover, more experienced with any reduction in compliance costs in monetary terms. firms using tax software pay less for compliance and such However, software usage was associated with a significant benefits decrease over time; experience with tax compliance reduction in time spent for tax compliance. This implies a in the first few years after the registration is very valuable, but reduction in effort using tax software is possible, but the firms learn incrementally less every year. reduction in cost is lost once tax staff salaries are considered. This finding in turn suggests the possibility of higher salaries being associated with the skill set requirements of the position Summary of research findings (for example, highly skilled accountants, who can use tax The study’s main results indicate that a reduction in tax software, command higher salaries). Firms reporting obstacles compliance costs associated with tax software usage is with “finding skilled workers” are subject to about a 5 to 7 indeed possible, but it is only likely to be realized in certain percent increase in tax compliance costs. cases. For instance, in the case of Nepal, estimates differed between negative 6 and negative 46 percent depending on In Armenia, the results show that tax software usage is how firms were grouped (for example, only firms that do all associated with a 47 to 63 percent increase in tax compliance tax compliance in house, only firms above the value-added costs (in local currency) and a 27 to 47 percent increase in tax threshold, only computer users, and so on). Yet, most of time spent to comply. However, these results were not robust the estimates are statistically insignificant and not very robust enough to allow for considering different sub-samples of to checks with the most relevant sub-samples. The statistically firms. More importantly, depending on “who undertakes significant estimates, although weakly significant, suggest a 30 tax compliance and uses the tax software,” firms experience to 46 percent reduction in compliance costs may be associated net gains or losses from using software. In general, certified with tax software usage in Nepal. accountants with graduate degrees using tax software spend less time for tax accounting then non-certified accountants Results with respect to Armenia are more pronounced because spend. Yet, in terms of the monetary cost of tax accounting the data provides more detail. The survey contains information incurred by firms, certified accountants with graduate degrees on the occupation and education of the “most knowledgeable using tax software cost more than non-certified accountants person” on tax compliance who also did the tax compliance with the same degree using software. This may be attributed work in the relevant fiscal year covered by the survey. to their experience and education with tax compliance as well This allowed the analysis to infer the extent to which the as certification that in turn render higher payments to the productivity of the person who uses tax software can improve highly educated, certified accountants. the effectiveness of the software in reducing compliance costs. Can Specialized Tax Software Reduce Tax Compliance Costs in Developing Countries? 3 Furthermore, after utilizing all this information, tax software could imply somewhat lower revenues for the government). It is usage was associated with a 47 to 63 percent increase in tax plausible that the use of tax software might help reduce errors compliance costs (in monetary terms, taking into account (to the benefit of both taxpayer and government), but the survey wage rates) and a 27 percent increase in time spent to comply. data did not allow that hypothesis to be verified. However, firms that hire accountants (certified or non-certified) with a high level of education (such as a graduate degree) Notes can mostly negate the increase in compliance costs that is otherwise associated with the use of tax software. For instance, See also U.S. IRS 2005 1040 Instruction Book, p. 79, showing time and cost 1 for preparing a Form 1040 including Schedule A and other schedules (but in firms that have a non-certified accountant with a graduate excluding Schedule D): self-prepared without software –16.7 hours; self- degree, using tax software can actually recover all the increase prepared with software – 22.7 hours. in compliance costs associated with tax software usage. http://turbotax.intuit.com/small-business-taxes. 2 The clearest evidence of reductions in tax compliance costs through use of 3 Thus, the net effect of using software on compliance costs in tax software was seen in Armenia among businesses that had skilled and monetary terms for such a firm will be zero relative to firms well-qualified accountants in charge of tax compliance, who were probably less prone to “over-using” the software. where the owner (or other non–accountant staff) does the tax compliance work. When compliance costs in time are References considered, the results are slightly different. In firms employing a certified tax accountant with a graduate degree, the use of Coolidge, J. 2012. “Findings of Tax Compliance Cost Surveys in tax software can actually save 12 to 24 percent of compliance Developing Countries,” e-Journal of Tax Research, Vol. 10, costs (measured in staff-time spent to comply). This figure no. 2, Oct. 2012, 250–79. for non-certified accountants is between 11 and 21 percent DeLuca, Donald, John Guyton, Wu-Lang Lee, John O’Hare, and (that is, slightly less than for certified accountants). These Scott Stilmar. 2007. “Estimates of U.S. Federal Income Tax Compliance Burden for Small Business.” 100th Annual results highlight the importance of the effective usage of tax Conference on Taxation, National Tax Association. software in realizing compliance cost savings for end-users. Guyton, John L., Michael P. Stavrianos, and Eric J. Toder. 2003. It is possible that non-accountants might be more likely to “Estimating the Compliance Cost of the U.S. Individual Income waste time fruitlessly running many different “scenarios” (for Tax.” 2003 National Tax Association Spring Symposium. instance, shifting expense categories or timing of realization of IFC (International Finance Corporation). 2011. The Costs of Tax losses) in an attempt to reduce the overall tax bill. By contrast, Compliance in Armenia. Report. Washington, DC. knowledgeable and experienced accountants might be more Yılmaz and Coolidge. 2013. “Can Specialized Tax Software Reduce Tax likely to make efficient use of tax software and avoid the Compliance Costs for Small Businesses in Developing Countries?” temptation to waste time with it. Informal paper. World Bank Group, Washington, DC. Overall, the study’s estimates show that tax software usage is associated with a reduction in compliance costs in certain _______________ specific cases. Yet, the estimates are usually not robust enough to control for different sub-groups of firms. This suggests that Analysis and note were prepared by Jacqueline Coolidge and Fatih a reduction may be expected for certain types of firms, but Yılmaz (Investment Climate Department, World Bank Group). probably not for all. The findings and views published are those of the authors and should not be attributed to IFC, the World Bank, the Multilateral Investment Policy implications Guarantee Agency (MIGA), or any other affiliated organizations. Nor Given that specialized tax software is usually a commercial do any of the conclusions represent official policy of the World Bank or product in most countries (unlike the e-filing software provided of its Executive Directors or the countries they represent. by the tax authorities in many countries), there is little to offer in relevant policy advice. Also, the ambiguous findings with regard The Investment Climate Department of the World Bank Group to the effect of tax software usage on tax compliance costs make helps governments implement reforms to improve their business it unclear whether policy should encourage the use of software environments and encourage and retain investment, thus fostering or not. It appears possible that an increase in compliance costs competitive markets, growth, and job creation. Funding is provided by associated with taxpayers’ use of software is balanced by the the World Bank Group (IFC, the World Bank, and MIGA) and over 15 perceived likelihood of a reduction in the tax bill (which in turn donor partners working through the multidonor FIAS platform. Contact Rajul Awasthi l Senior Private Sector Development Specialist l Investment Climate Email: rawasthi@worldbank.org TEL: 1-202-473-1229 www.wbginvestmentclimate.org