Posted by William B. Trautman*
The recent financial crisis has led not only to significant public budget deficits in the United States and around the world, but also to a heated debate over what to do about them. Some argue for increased government spending to fund stimulus programs while others argue for reduced spending. Little attention has been paid, however, to increasing the collection of tax that is already owed. While tax administration information technology projects have a long and spotty track record, the advent of electronic filing in XML format and related technologies has significant potential for improving the collection of tax revenue, particularly with respect to complex business enterprises and high-wealth individuals.
This post will discuss an XML-based technology that the US Internal Revenue Service (IRS) has tested on the largest U.S. corporate taxpayers over a recent 18-month period. The technology has yielded cost savings in terms of more streamlined audit planning and has identified significant corporate tax revenue that otherwise would have gone undetected. The technology is essentially a data management and analysis system that allows one to construct virtual economic enterprises with a degree of detail constrained only by the available data and estimate the tax compliance risk of those enterprises in real time.
Tax authorities have long struggled with assessing the accuracy of income reported by complex taxpayers for the purpose of selecting tax returns for audit. The infinite number of organizational structures associated with the population of taxpayers whose business activities involve multiple entities, tiers, and ownership interests is effectively impossible to represent in two-dimensional (e.g., relational) data structures used by standard analytic software. The result is that tax authorities simplify the structure of the information for storage and analysis, but the associated loss of information may have a detrimental effect on the richness and quality of the risk assessment.
XML and object-oriented programming technologies have made it possible to overcome this limitation and model complex taxpayers in a high degree of detail in real time. It is possible to read in information in XML format and construct a virtual enterprise out of computer objects that represent the basic building blocks (e.g., subsidiaries or partnerships) of complex enterprises. These objects can contain information from tax returns, separately filed related returns, or financial statements, and they can be structured in a way that captures the organizational structure of enterprises. This allows tax authorities to search for certain structures in the context of the virtual enterprises that may be indicative of tax compliance risk, such as multi-tiered flow-through structures that may be associated with tax shelters.
A prototype version of this software, known as Spyderware, was developed using open-source XML and object-oriented technologies and other publicly available information. Spyderware is written in Java, is platform independent, and runs on any computer with the Java Runtime Environment (JRE) installed. During IRS testing, the software was installed on an IRS laptop with a connection to a back-end database housing XML tax return documents and financial information, and it applied the IRS’s proprietary international risk measures to over 6,000 of the largest corporate tax returns. This suggests that the software can be installed and run with minimal investment in hardware, software or time.
--------------------------------------------------------------------------------
* William B. Trautman (wbt@taxCrest.com) is a Senior Economist at the IRS and President of taxCrest, LLC, an entity established for further development and marketing of XML-based tax administration software. The views expressed in this post do not necessarily reflect the views of the IRS, and the IRS does not endorse the software.
Note: The posts on the IMF PFM Blog should not be reported as representing the views of the IMF. The views expressed are those of the authors and do not necessarily represent those of the IMF or IMF policy.
