Customs Tariff Classification and the Use of Assistive Technologies

Customs Tariff Classification and the Use of Assistive Technologies

This paper aims to help inform discussions about customs tariff classification and the use of assistive technologies, especially those found online. It offers a broad background overview of classification practices and requirements by reference to the World Customs Organization's Harmonized System, incurred challenges and costs, and the types of technologies deployed to assist users in their classification endeavors. Reviewed assistive technologies and solutions range from information tools that make it easier for users to find classification-relevant information (such as rulings and guidance notes) to Artificial Intelligence approaches that aim to suggest the most appropriate codes - with varying levels of success. Suggestions with examples about how to evaluate assistive classification technologies are provided. It is argued that there is considerable potential for such technologies, however, this inevitably depends on how well they can understand HS-relevant characteristics, the quality of data that supports them, and how well they have been developed or trained. The prize includes better compliance outcomes and improved trade facilitation by reducing classification-related costs. However, it is essential that policy makers and users understand the opportunities and limits of assistive classification technologies.

Introduction

It is hard to imagine customs control arrangements without commodity codes. This is because there are all sorts of goods with differing characteristics - including shape, form, use, composition, and production stage. Categorization into codes helps make their control and reporting more manageable. Some practitioners may even refer to commodity codes as 'buckets' into which goods with similar characteristics can be sorted - classified - and be made subject to a trade tariff that is specific to the assigned code. In over two hundred countries customs commodity codes are now based on the World Customs Organization's Harmonized Commodity Description and Coding System - commonly referred to as the Harmonized System. HS codes also serve as the starting point for preferential origin rules, quantitative restrictions, and tariff quotas. Often, HS-based customs codes are referenced in other regulatory domains, too. This might be in the collection of import Value-Added Tax and Excise Duties, Sanitary and Phytosanitary type controls, trade sanctions, prohibitions, and restrictions, and many other trade- and cross-border related control measures. HS codes also play an important role in the collation and analysis of international trade statistics.

But despite the significance of HS codes in the administration of the cross-border trade and customs environment, determining the correct HS classification is often considered to be difficult. Expert training is usually considered to be necessary. And although strict classification rules apply (described later in Box one), errors are frequently made. The Auditor General of Canada estimates that twenty percent of goods brought into Canada during the fiscal year twenty fifteen-sixteen was misclassified. It is no secret, evidenced by many tribunal and court rulings - not to mention the discussions among classification professionals - that classification rules and code descriptions can be interpreted differently. Most customs practitioners can quickly list many examples for which tariff classification is difficult.

But despite the significance of HS codes in the administration of the cross-border trade and customs environment, determining the correct HS classification is often considered to be difficult. Expert training is usually considered to be necessary (European Commission, 2019b; WCO, 2019). And although strict classification rules apply (described later in Box 1), errors are frequently made. The Auditor General of Canada (AGoC), for example, estimates that 20 per cent of goods brought into Canada during the fiscal year 2015-16 was misclassified (AGoC, 2017). It is no secret, evidenced by many tribunal and court rulings - not to mention the discussions among classification professionals - that classification rules and code descriptions can be interpreted differently. Most customs practitioners can quickly list many examples for which tariff classification is difficult (Table 1).

Given such HS classification challenges, there is a strong case for policy makers to routinely make assistive technologies available to users, and thus help ease their classification burden. Such tools have relevance for the entry and validation of customs codes into electronic customs systems as well as wider single window environments. There are also many online services that require users to be familiar with HS codes. And there are many types of resources - especially online - that seek to reduce classification-related challenges. Recently much is also made of Artificial Intelligence - especially where parallels are drawn to innovations elsewhere, such as in biotech developments, financial and legal services.

Specific to the trade and customs domain, AI applications are already being explored within the context of strategic goods controls and also discussed within the context of customs tariff classification. Increasingly, regulators are also asked to ensure that new technologies are used responsibly. This paper reviews current tariff classification practices and the increasing use of new, assistive technologies - especially those found online. The underlying motivation is to help inform discussions among users, developers, and policy makers about tariff classification requirements, practices, and the responsible use of tariff classification technologies.