Submission

Automated Audit Criteria Selection System (ACSS)

ACSS is an analytical tool for the tax administration of Tunisia to select taxpayers for audit. It allows efficiency and transparency.

1. Zouheir Bouchaddakh
2. Arlette Ngo Badjeck
3. Azza Chemkhi
4. Stephen Wilcox

1. Tax Administration Specialist
2. Project & Communications Specialist
3. Technical Interpreter/ Translator
4. Tax Administration Expert

1. zbouchadakh@tunisiatcp.org
2. anbadjeck@tunisiatcp.org
3. achemkhi@tunisiatcp.org
4. shw042@darimmir.com

Middle East

(AMEG) Tunisia Tax and Customs Reform Pilot Activity (TCP)

Democracy and Governance

TCP is assisting the Tunisian tax administration to modernize its procedures by making use of electronic technology to increase efficiency, create transparency and demonstrate integrity. The Tunisian tax system is based on voluntary compliance. Taxpayers make payments based on the returns they complete. The returns are accepted in the first instance without scrutiny. However, the tax administration may follow up and review a tax position in depth with a tax audit. Given scarce human resources, the tax administration can verify only a very limited number of taxpayers' returns. The problem is: how does the tax administration choose what returns to review in the most transparent, fair, and efficient manner? ACSS is an innovative tool that helps solve this problem – it is based on scientific criteria to best target and select returns to verify that taxpayers comply with tax laws. Being an automated system, it creates objectivity by removing human bias.

Tunisia is undergoing a democratic transition to free itself from the bad governance that characterized the regime prior to the 2011 revolution. That regime controlled and silenced political opponents by targeting their personal and business activity through tax audits. Tax staff were thought to be corrupt. In contrast, ACSS applies objective means to select cases, enabling the Tunisian tax administration to improve its image vis-à-vis taxpayers with a more transparent and fair approach to auditing. Moreover, establishing objective criteria for the selection of audit cases detaches a tax administration employee from the selection process. This helps diminish any allegation of corruption or unfairness. ACSS is also unique because it allows more cases to be selected for audit and in a shorter period of time. In sum, implementation of ACSS contributes to greater tax fairness, efficiency, and helps the authorities to encourage voluntary compliance.

Automated Audit Criteria Selection System (ACSS)

Tunisia

Big Data

02/2016-04/2016

ACSS was designed in house by the Tunisian tax administration in partnership with TCP, using SQL and Oracle language. ACSS assigns a risk score to all taxpayers. The higher the risk score, the more likely that the taxpayer is underreporting their tax liability. The risk score is based on multiple criteria which look at different sources of information. There are three types of criteria used.
1. Comparison against peers.
2. Logical test.
3. Ratio test.
Taxpayers are classified into groups using type of activity, entity and turnover to calculate the national average.
The criteria takes information from multiple sources available to the tax administration:
1. Annual income tax returns
2. Aggregated monthly VAT returns
3. Aggregated tax payment information
4. Customs data
5. Social Security information
6. Municipality registration of sales contracts
7. Tax audit information.
Using this information, a total score is given to every taxpayer.

ACSS has enhanced the ability of the Tunisian tax administration to prioritize the audit of the riskiest taxpayers. TCP supported the tax administration to roll out ACSS via eight regional offices throughout Tunisia. Several regional offices recognized the immediate increase in efficiency that ACSS produced and noted that ACSS was yielding cases of non-compliance that would not have been detected using traditional case selection methods. Work is ongoing, with some cases expected to yield large results.
In sum, the ACSS increases the efficiency in which the tax administration selects cases for audit, providing information for timely decision making to verify that taxpayers comply with tax laws. In addition, the objective criteria integrated in the system increases the integrity and the transparency of case selection. This will strengthen taxpayer trust and confidence in tax administration, which positively affects rates of voluntary compliance.

Through the pilot, bugs were detected that produced incorrect scores. In addition, some criteria did not produce the expected score due to poor data quality. In response, the bugs were corrected and changes were made to the computer logic to address poor data quality.
A review of the hardware needs of the ACSS team showed that team members needed more powerful equipment to run the system given that it had to process hundreds of thousands of records. ACSS was installed and laptops delivered.

ACSS is flexible and could be replicated for tax systems in other countries. The criteria may be modified to adapt to the taxation situation of any country. ACSS could be used where risk management is applied or for selecting cases according to well-defined criteria. For example, in order to select the “right” candidates for a vacant job, an automated selection criteria will be useful in the examination of thousands of job applications.

Automated Audit Criteria Selection System ACSS.pdf