One impact of digitalisation has been an increase in small parcel trade. As the value of parcels often falls below the de-minimis thresholds adopted by customs authorities, there is a concern – albeit one that recognises that the impact on overall values of trade is likely to be marginal – that small parcel trade may not be fully captured in official statistics. Significant improvements already underway in customs clearance procedures and tracking systems in many countries will help to establish whether there is systematic underestimation.
More significant challenges exist in the area of (digitally delivered) trade in services, particularly to households. Many European economies are now beginning to use VAT returns from firms to improve on current measurement. These approaches typically lead to upward revisions at the product level. For example, households import 6% and 30%, respectively, of total imports of computer services and audio-visual products in Denmark, but the overall impact remains small, amounting to revisions of less than 0.4% of total imports.1
Ensuring that cross-border flows of intellectual property-related services align with core accounting concepts remains a significant challenge. Even when mismeasurement is not an issue, there remain challenges around interpretation2
, as was illustrated by the 26% upward revision to Irish GDP in 2015 (OECD, 2016). The broader issue of measuring intra-firm trade is exacerbated by large non-monetary data flows and delivery of services via affiliates abroad, which are also difficult to capture.
Notwithstanding these issues, a key problem for the development of statistics on digital trade is that current statistical classification systems do not routinely delineate digitally ordered or delivered trade flows from those that are not. In other words, it is hard to identify digital trade through the prism of current classifications.
To address these challenges, countries are exploring new data sources, such as credit card information, and developing projects linking business register data with customs data to provide information on the size of imports and exports by e-tailers (classified as NACE 47.91), or linking other sources. They are also exploring the scope for adding new questions to existing surveys. Costa Rica, with support from UCTAD, recently developed estimates of digitally delivered services using this approach. However, resource constraints and pressure to reduce respondent burden present a challenge in many countries.
Other challenges relate to when, how and by whom trade flows should be recorded. Digital intermediary platforms, which facilitate transactions for a fee, do so without ever taking ownership of the products involved. The identification of these platforms in business registers, their classification in terms of the actual services they provide, and the treatment of the transactions they facilitate – including which parts should actually recorded as being cross-border, and with which partner country – pose significant conceptual and empirical challenges.
Finally, current frameworks also struggle to identify the take-up of digital tools and technologies to engage in trade. OECD’s Statistics and Data Directorate Informal Advisory Group on Measuring GDP in a Digitalising Economy is conducting work to address this need (see also page 2.11).
