Lessons learned: Incorporating data analytics in tax curricula

Editor: Annette Nellen, Esq., CPA, CGMA

Data analytics is a valuable tool that is used to gain insights from ever-increasing volumes of data. To help students become familiar with how to apply this method of analysis to tax data, educators should teach data analytics in the classroom. They can do so by mirroring a process currently being used by entry-level accounting professionals called “extract, transform, and load” (ETL) and focusing on developing students’ data analysis and visualization skills. These educational experiences will give students a better foundation for a successful career in tax, no matter what career path in the profession they may pursue.

While it is important to incorporate analytics into the tax curriculum, the implementation may seem daunting. To help, this column identifies ways that educators can overcome four major challenges: (1) What specific tools should be taught?; (2) What data analytics processes should be taught?; (3) How can data analytics be incorporated into a course that is already packed?; and (4) How can data analytics cases that are already available be adapted to meet specific needs?

Along the way, this column highlights new tax data analytics cases that can enhance students’ learning experiences.

What specific tools should be taught?

Should you teach A2019, Access, Alteryx, Azure, Blue Prism, Java, C++, Excel, Gephi, Knime, Power BI, Python, R, SAS, SQL, Tableau, and/or UiPath? The anticlimactic answer is that the specific tool you choose to teach is less important than the data analytics process. However, before discussing the data analytics process, this column offers some practical advice for selecting data analytics tools.

First, teach the tools that are used by the firms that hire your students. Your program’s advisory board members, firm recruiters, and recent graduates can provide valuable insight as to the needs of your particular job market. It may be useful to inquire if there are tooling transitions underway, particularly with the increasing use of cloud-based analytical applications.

Second, do not underestimate the importance of fundamental Excel skills. Excel remains a staple for employers, partially because clients, managers, and partners commonly feel comfortable with Excel. However, employer feedback continues to indicate that students’ Excel skills fall short of expectations. Since the student population is diverse, ranging from tech-dependent to tech-savvy, instructors must be mindful of covering the basics (e.g., file management, copy and paste, formulas, cell labeling, formatting), in addition to identifying projects that allow students to develop advanced skills (e.g., VLOOKUP, conditional formatting, PivotTables, and visualizations). One case that may help with Excel skills is Christine Cheng, Pradeep Sapkota, and Amy J.N. Yurko’s “A Case Study of Effective Tax Rates Using Data Analytics,” a forthcoming article in Issues in Accounting Education and the winner of the 2019 ATA/Deloitte Teaching Innovation Award. The casefocuses on developing students’ data evaluation, PivotTable, and data visualization skills using Excel. The support material for this case includes custom instructional videos that address basic and more advanced Excel skills.

Educators who are ready to introduce a more powerful data analytics tool can quickly move students from Excel to Power BI, since the functions used in Power BI are similar to Excel’s. Using an add-on approach (such as beginning with Excel and then introducing Power BI) can ensure that you educate students at all levels along the technology spectrum while helping students understand that they can adapt skills learned in one tool to other tools.

Third, while Excel and Power BI are powerful, these tools are not universally ideal for data analytics. In response to the need for more advanced tools, Alteryx (built around the ETL process) and Tableau (designed for effective data visualization) are increasing in importance in the accounting profession. Further, there is continued convergence among these tools in an effort to become the preferred data analytics tool (for example, Tableau now offers data transformation capabilities, and Alteryx now offers data visualization capabilities). Both programs offer cost-free licenses to students and faculty. The low-code characteristic of these tools is particularly appealing to students. Alteryx’s repeatable workflow and ability to track data transformations throughout the ETL process is particularly appealing to employers. Alteryx or other Windows-based data analytics tools can be run on iOS-based operating systems if the student uses VMware, such as Boot Camp.

Lastly, instructors should consider incorporating at least an introduction to a code-based tool, for example, Python, R, or SAS. Code-based software programs are powerful tools that provide comprehensive data cleaning, manipulation, and analysis capabilities. Much of the stress can be reduced by providing students the appropriate code and incorporating class demonstrations. An introduction will not make students fluent in coding, but it may help students overcome their aversion to code-based analysis and be the spark that motivates further development. In introducing coding, the authors’ recommendation based on experience is to avoid providing students code in a format that is easy to copy and paste. It is important that the students key in the code to help them learn the process.

What data analytics processes should be taught?

Tax and accounting professionals typically use the ETL process to evaluate data. Readers can learn more about the ETL process from James Zhang, Snigdha Porwal, and Tim V. Eaton’s article, “Data Preparation for CPAs: Extract, Transform, and Load,” 230-6 Journal of Accountancy 50 (December 2020).

Efficient data analysis has a purpose. Before tax professionals can begin the ETL process, they need to have a good question in mind. There are varied approaches to teaching students how to ask a good question. Cases, for example, mimic the process where a client or senior colleague might give an associate a task to investigate. Faculty can also work on advanced critical thinking by having students identify their own research questions to investigate. However, the latter may require that the faculty act as a sounding board, especially if students have limited experience in critical thinking.

The process of extracting data

Once the student has a research question, he or she must extract the appropriate data. Fortunately, there is ample data, particularly in tax, thanks to the plethora of data published by the IRS and state and local governments. For example, Cheng and Anu Varadharajan’s “Using Data Analytics to Evaluate Policy Implications of Migration Patterns: Application for Analytics, AIS, and Tax Classes,” a forthcoming article in Issues in Accounting Education, presents a case that relies on the migration data from the IRS Statistics of Income department (available at www.irs.gov).

Faculty can also rely on data obtained from data aggregators, like Compustat or directly from the SEC’s XBRL files. Cheng et al.’s “A Case Study of Effective Tax Rates Using Data Analytics,” which focuses on effective tax rates (ETRs), has students use either Compustat data or the SEC’s XBRL data available at www.sec.gov. While there is an initial learning curve required for using XBRL data, the case study describes the process of gathering tax information using the XBRL tags. Students exposed to the use of XBRL tags will have several valuable opportunities. First, they can gather data in an “as reported” format and better learn the process for financial reporting for tax expenses. Additionally, recent research by Casey Schwab, Bridget Stomberg, and Junwei Xia, “How Well Do Effective Tax Rates Capture Tax Avoidance?” (2020) (working paper available at www.semanticscholar.org), found that SEC data is more accurate than Compustat for important information like tax footnotes. Finally, instructors can use information published by XBRL US to highlight issues related to data quality by having students consider how the validation rules help ensure data quality (visit xbrl.us) and investigate recent filing results and quality checks (visit xbrl.us).

There are distinct data benefits in both “A Case Study of Effective Tax Rates Using Data Analytics” and “Using Data Analytics to Evaluate Policy Implications of Migration Patterns: Application for Analytics, AIS, and Tax Classes” cases, which instructors should consider when selecting a case. One, the cases do not require that the students work with extremely large datasets. An extremely large dataset may create problems for students who lack access to a quality computer and/or reliable internet for remote access or cloud-based applications. You do not need “big data” to teach analytics because the tools and ETL process are the same regardless of the size of the dataset. Two, these cases provide options for free data. Three, the data for these cases changes every year. Therefore, you can reassign the case yearly, and students will get results that differ from the prior year’s. Four, the data can be divided so that different students or teams evaluate different data. For the ETR (migration) project, students can be assigned different industries (states). If you develop your own case, these are some of the data factors that you must consider.

You may also obtain data by asking alumni if they would be willing to provide data or to check fictitious data you create for authenticity. Using data provided or validated by alumni comes with a potential additional benefit of promoting real-world interactions between students and businesses.

The process of transforming the data

After extracting the data, students must transform the data to prepare for evaluation. Data analytics tools continue to advance to deal with unstructured data. However, data transformation will remain a vital part of the ETL process because data is typically not collected or stored perfectly, without any blanks or miscoding issues. Additionally, data transformation can be required because of the lag between the creation and use of the data. The current pandemic provides a perfect example of this. Who would have anticipated that critical information regarding individual location preferences, data that has been collected and reported electronically since the early 1990s, would be important to state and local governments as they project budget shifts arising from the rapid shift to remote-work arrangements during the COVID-19 pandemic? This location preference data must be transformed and potentially combined with other, imperfect sources of data to provide the best insights in the current business environment. The old adage still rules — “garbage in, garbage out” — so quality transformation is essential.

How can data analytics be incorporated into a course that is already packed?

Tax professionals who have an intimate knowledge of the complex interactions among taxpayers, tax law, and regulators are best positioned to perform the most important parts of data analytics, including asking the right questions, identifying relevant data, understanding how to transform the data to answer current questions, and completing the ETL process by loading the transformed data into the appropriate programs for predictive analytics or for creating effective visualizations. The quality of the analysis for decision-making is intimately tied to the capacity for critical thinking and problem-solving. Fortunately, the latest tools put data analytics into the hands of a group of individuals extremely skilled at critical thinking: tax professionals.

Thus, tax courses should continue to focus on teaching tax first. With major tax reforms and shifting business environments, there is never a shortage of material to cover in tax courses. But it is important to recognize the opportunities afforded by incorporating data analytics tools and processes into tax classes. Teaching the ETL process develops the analytics mindset and analysis skills. Linking tax accounting to analytics develops students’ understanding of the core tax concepts explored in the analytics case. A well-designed analytics case that explores an interesting research question will develop students’ research, critical-thinking, and communication skills. Further, an interesting tax question will engage students and show them how fascinating tax accounting can be as a profession.

One of the best tips that can be provided for incorporating data analytics into your tax course is to focus on a topic that has always been challenging for students to understand. Several recent tax data analytics cases can help you get started.

Examples of tax data analytics cases

Faye Borthick and Lucia Smeal’s case, “Data Analytics in Tax Research: Analyzing Worker Agreements and Compensation Data to Distinguish Between Independent Contractors and Employees Using IRS Factors,” 35 Issues in Accounting Education 1-23 (2020), requires students to use Access to evaluate the IRS facts-and-circumstances tests for whether workers should be classified as independent contractors or employees. In the gig economy, where students often drive for ride shares, for instance, this case can help students gain a better understanding of what facts and circumstances mean while connecting to a topic they are already familiar with.

Cheng et al.’s “A Case Study of Effective Tax Rates Using Data Analytics” is an excellent resource to help students learn about ETRs. This is always an important topic for businesses, but it is particularly important at this time of remote working, when businesses may be evaluating the tax laws of alternative jurisdictions. This case is particularly helpful in a corporate tax class because the evaluation firms with outlier ETRs develop students’ understanding of book-tax differences. This case primarily uses Excel, but it also includes instructions that introduce Tableau for data visualizations and Alteryx to run predictive analytics so that students can use linear regressions to explore factors that are associated with ETRs.

A basic component of tax planning is to understand how taxes can influence taxpayer decisions. The migration case by Cheng and Varadharajan mentioned earlier can help students see this, by investigating whether taxpayer location decisions are driven by tax law changes. This case allows students to learn the top 10 most used tools in Alteryx to engage in the ETL process and in Tableau to visualize the results.

A related case by Roby Sawyers, Tom Dow, and Lynn Jones titled “Tax Reform: A Case Using Data Analytics,” published by Van-Griner Learning in 2020, helps highlight that congressional leaders are concerned about the impact of new tax provisions on their constituents. This case requires students to use Excel and Tableau to analyze how the new state and local tax deduction limitation imposed by the law known as the Tax Cuts and Jobs Act, P.L. 115-97, may affect taxpayers within a metropolitan statistical area.

There are also several cases that help highlight the influence of taxes on small businesses. Cheng, John Eagan, and Yurko’s “ChicagoLand Popcorn — Examining Online Retailer Nexus Following Wayfair Using Data Visualization and Robotics Process Automation” (2020), available from the authors upon request, provides students with the opportunity to combine legal tax research with data analytics to help students understand how tax reform may affect small online retailer expansion decisions. In this case, students researched state-specific economic nexus requirements using online tax research tools such as CCH and RIA and used either Tableau exclusively or Automation Anywhere’s A2019 robotics process automation tool along with Tableau to evaluate whether an online retailer had established economic nexus in particular states, and the subsequent tax consequences.

Lauren Cooper, Kimberly Key, and Mollie Mathis’s “S Corporations and IRC Section 199A: Incorporating Excel Into Tax Planning Scenarios,” a forthcoming article in Issues in Accounting Education, discusses how Excel can be used to help students understand the intersection of S corporations and the Sec. 199A deduction. Since the new qualified business income deduction rules are complex, this case may provide an ideal resource for helping students gain a better understanding of the rules and the tax implications of these rules in a small business decision-making context.

Using the cases

Several of these cases come with short (5- to 10-minute) videos that demonstrate the data analytics skills. For example, you can check out Cheng’s YouTube channel at www.youtube.com. Her custom videos walk faculty and students through the data skills that are required in each of the tools to complete the data analytics cases she co-authored.

It is important to determine how much time to allocate to introducing data analytics into your tax course. The authors have found that the best approach is to provide some in-class time focusing on the tax technical and critical-thinking components of the case. For those instructors who want to ensure individual students are completing their own project, consider requiring students to add their name to a variable. This time does not take away from what you are already teaching, since it is a different approach to teaching the same tax technical process. However, you may also want to allocate some time to demonstrating the tools to help students become comfortable with them.

The authors caution against spending too much time in class on the tools demonstration. It may take too much class time, and the live demonstrations will inevitably be too slow for some students and too fast for others. The authors recommend referring students to videos already created and/or creating your own custom videos. The videos mitigate the problems that arise from having a varied student base, allowing students who are reluctant or unfamiliar to watch the video multiple times, while allowing students who are familiar to skip the videos for steps they can confidently complete.

The authors also recommend that you think about which case might be best introduced in an accounting information systems course, or an introductory tools course. This suggestion serves several purposes. First, in order to help students understand tax careers, your program should introduce a case that demonstrates how data analytics can be used to solve interesting tax problems as early as possible. Second, students benefit from seeing the data analytics tools multiple times to help them maximize their learning and comfort level in working with data analytics tools. Finally, the cases help you work on building the most important skills that will be required of students if they enter careers as tax data analytics professionals: critical thinking, problem-solving, and communication.

How can data analytics cases that are already available be adapted to meet specific needs?

It may be the case that none of the data analytics cases highlighted in this column align with topics that you already cover. You might consider adapting an existing case to fit your needs, instead of trying to reinvent the wheel completely. Your adaptations of a case can focus on tailoring the tax topics and critical-thinking components to those topics and level of critical thinking that are best suited to your class. For example, consider the migration case by Cheng and Varadharajan, which was developed prior to the COVID-19 pandemic. One adaptation is a consideration of how the pandemic might impact these migration patterns and what effect that could potentially have on state or local tax revenues. While the data for 2020 is not yet available, students can evaluate whether current media reports of migration trends are similar to trends that occurred pre-pandemic and/or use predictive analytics to project future trends. Predictive analytics could also be used to evaluate how migration trends are associated with state tax revenues and use these results to predict the expected impact of migration trends occurring in the present on state and local tax revenues. Future students could then compare the actual data, when it becomes available, to the predictions made by prior classes.

Another adaptation of this case is to have students focus on tax liability. For example, adapt the case to an internal audit perspective, where companies are trying to figure out their exposure to income tax reporting requirements. Alternatively, adapt the case to a merger-and-acquisition due diligence perspective where an acquirer is trying to figure out the potential unaccounted tax liability that has developed due to the targets’ remote-work arrangements. Finally, this perspective could stem from a state tax audit perspective, either just based on confirming amounts reported for salary apportionment in light of remote-work arrangements or evaluating employment requirements that are often tied to special tax incentives provided to corporations.

There are other opportunities to expose students to data analytics and manipulation in the area of tax compliance. Often, tax practitioners upload client trial balances into tax software for the preparation of partnership or corporate tax returns. This data needs to be coded so the tax software can put each trial balance item on the correct return line. These codes are sometimes referred to as tax return codes (TRCs). The assigning of TRCs can be very time-consuming and repetitive.

Cheng, Shaw, and Luke Watson are in the process of developing a case study that provides students with a trial balance that needs to be coded with TRCs using Alteryx. This project could be expanded to illustrate the benefits of ETL by requiring them to code multiple trial balances. Students would quickly realize the benefits of data tools when they see how much work can be saved over a manual process. Another benefit of this project is that it can be incorporated into an existing tax compliance project. The coded trial balance can be used with tax preparation software to complete a tax return project. The importance of this project as a teaching case that mimics real-world professional experience for tax compliance is underscored by the announcement of a partnership between Alteryx and Thomson Reuters this past summer (see tax.thomsonreuters.com). The authors expect that the case will be ready for limited distribution by the end of the spring 2021 semester.

While these are just a few examples, hopefully, this helps spur thoughts about how you can adapt the tax topic areas and critical-thinking components to best suit your course. If not, continue to monitor the education journals, as there are several additional tax data analytics cases on the horizon. Whatever case you select, the authors strongly recommend that you complete the project yourself from start to finish. This will develop your technical skills, prepare you for student questions, and potentially inspire ideas for future analytics projects.

The Securities and Exchange Commission disclaims responsibility for any private publication or statement of any SEC employee or commissioner.This article expresses the author’s (Christine Cheng’s) views and does not necessarily reflect those of the commission, the commissioners, or other members of the staff.

Editor’s note: The authors appreciate helpful comments from Julie Marlowe and Mike Willis.   

Contributors

Annette Nellen, Esq., CPA, CGMA, is a professor in the Department of Accounting and Finance at San José State University in San José, Calif. Christine Cheng, Ph.D., is an assistant professor of accountancy and a visiting scholar with the SEC’s Office of Structured Disclosure, and J.R. Shaw, Ph.D., is an associate professor of accountancy, both at the University of Mississippi in Oxford, Miss. Amy J.N. Yurko, Ph.D., is an associate professor of accounting at Duquesne University in Pittsburgh. To comment on this article or to suggest an idea for another article, contact thetaxadviser@aicpa.org.