In the first part of this article [1], we addressed the importance of integrating Tax Administration systems with the rest of the tax ecosystem to achieve “Tax Administration 3.0”, mentioning that functional challenges are one of the three main challenges of this process. In this second part, we will focus on these challenges.

Definition of a common language

As stated in the practical guide for data governance in Tax Administrations [2], an excellent reference provided by CIAT, it is central to build a business vocabulary, consisting of data names and standardized definitions, documented in a glossary. Its purpose is to ensure consistency and mutual understanding of tax terms, a fundamental issue for the tax ecosystem.

For example, when integrating systems, it is crucial to define whether data representing the amount of an invoice appears in local currency or in the issuing one. Having these clear definitions offers multiple benefits, among which are:

• Avoiding misunderstandings, based on a shared vision about the data.
• This optimizes times, by avoiding unnecessary consultations and corrections.
• This unifies the information received by providing common sense.

Although we may detail these definitions in the exchange documents, in other cases it is not so easy to identify them. As Dr. Peter Aiken points out in [3], in addition to having these definitions, it is vital that they are known by those who use them. For example, what are the conditions for a company to qualify as small and where can you get the list of those that meet that criterion? This information could become available to the ecosystem through a simple website.

Another important aspect of this “common vocabulary” is the reference data, such as tables of countries or units of measurement. In these cases, a good practice is to use international standards, such as the ISO 3166 table (https://www.iso.org/iso-3166-country-codes.html). Taking advantage of available solutions for common problems is part of what institutions should consider for the 3.0 model.

In addition to reference data, there is master data. According to the DAMA dictionary [4], these are “the data that provide context to business activity in the form of common and abstract concepts”. In this line, they include the details of key objects such as customers, products, employees, and suppliers. In the case of Tax Administrations, data about taxpayers (name and surname, date of registration, address, etc.) are, without a doubt, the master data par excellence.

The State is responsible for managing both types of data, which poses the challenge of defining which agency is responsible for updating and maintaining them, and how other agencies can use them within the framework of their processes, in order to avoid taxpayers being responsible for providing them recurrently. For example, it would not be logical for two state agencies to use different country codes and/or to request the same information from taxpayers, independently, for each of their services. This situation not only undermines integration, but also leads to duplication of data, redundancy in processes, and in many cases to the generation of inconsistencies that affect the quality of information in the hands of the State.

Data quality

We say we need quality data, but what do we mean? An accepted definition is that quality is “the suitability of the data for the intended use”. This approach puts emphasis on a central aspect: we measure quality according to an objective, and we work on it to make better use of the information we have. It has to do with the level at which they meet expectations, that is, about what we expect from them, fundamentally in relation to the purpose for which we obtained them in the first instance.

In the context of systems integration, the quality of the data we send and receive is truly relevant, and the level of trust that exists around those exchanges. For example, if a tax administration receives data on taxes withheld from employees, is it reasonable for the amount of taxes to exceed the value of the salary? Or what if deceased people are present in the payroll?

On the other hand, when sending information, it is essential to validate it. One way to do this is by measuring its quality through a set of metrics for certain attributes or dimensions (for example, number of cases, maximum, minimum amount, etc.) and checking that they maintain consistency over time.

In the interaction between taxpayers and the Tax Administration, the idea of continuous transaction control (CTC), as established by CIAT in [5], requires sending the data at the end of the commercial transaction. A recent IDB report [6] indicates that almost ninety countries implement electronic invoicing. In several of them, the Tax Administration uses this data to pre-fill the tax returns. If an invoice misstates the amount of VAT withheld, this harms both the sender and the receiver, resulting in erroneous declarations.

To avoid these situations, active data control is necessary that validates the information as soon as possible. Returning to the example of the invoice, the taxpayers are those who should strengthen the controls in their systems. If, however, the erroneous data reaches the Tax Administration, it should issue alerts to out-of-range income and reject the operation or require confirmation.

In this regard, I find interesting what OECD describes about the common errors in the CbC report [7]. This type of documentation is especially useful for the private sector and for countries which can use it to implement more robust controls in their systems.

We know that data has quality problems, how can we start? The ideal is to choose a sensitive topic that allows you to see quick results and keep in mind that quality is not a project, but a continuous control process, which requires the definition of metrics and constant evaluation to measure the level with which they meet expectations.  The actors involved share this same process, both in the technical and business sectors, and this begins long before the data collection process.

Conclusion

In this second part of the article, we address the functional challenges of systems integration, focusing on the place occupied by the construction of meaning around data and the generation of a common vocabulary that we can document, for example, in glossaries. Also, from this perspective, we highlight the relevance of data quality, as a continuous process that allows us to generate trust about information that we exchange and sustain it over time. In the third and final part, we will focus on the legal and data protection aspects, understanding the importance of making an ethical and responsible use of the data in the hands of the Tax Administrations, in accordance with current regulations and in a framework of respect for the rights of taxpayers.

References

29 total views, 29 views today