4. Conclusions, discussion and recommendations
This problem is compounded when trying to identify cross border transactions of digital products and services. Because in a digital economy ‘borders’ are often much less relevant and clear. While economically speaking this has had an enormous stimulating effect on cross border transactions it has also made identifying many of these transactions more and more difficult (information paradox). A change of ownership of a good or purchase of a service increasingly happens by a simple touch of a mouse button. But whether the transaction is digital is not always accurately reported.
Despite these difficulties it is possible to provide reasonable estimates for the activities explored in this research. Often this was done by combining available statistical data sources, augmented with external sources and further refining them by applying relevant assumptions.
In most cases the estimates can be based on already existing statistical data sources, providing a solid foundation to work with. In other cases, such a source simply is not available and we must therefore rely on external sources and sometimes more experimental innovative techniques to generate an estimation.
| Year | Goods via e-commerce as a sales channel (billion euros) | Database services and database originals (billion euros) | Dwelling services due to digital intermediation (billion euros) | Digital intermediation of other travel service (billion euros) | Digital intermediation of hotel services (billion euros) | Taxi services due to digital intermediation (billion euros) |
|---|---|---|---|---|---|---|
| 2010 | 3431 | 1596 | 0 | 23 | 17 | 0 |
| 2011 | 4163 | 1451 | 2 | 25 | 18 | 0 |
| 2012 | 4558 | 1790 | 16 | 28 | 21 | 0 |
| 2013 | 4130 | 2821 | 46 | 27 | 20 | 0 |
| 2014 | 5272 | 2952 | 76 | 28 | 21 | 1 |
| 2015 | 4916 | 2850 | 169 | 29 | 21 | 2 |
| 2016 | 4747 | 2967 | 341 | 29 | 22 | 7 |
The import of all activities is shown in graph 4.1. Contrary to exports, imports of digitised products do show a year-on-year growth, but not as exponential. The import of dwelling services due to digital intermediation and the import of taxi services due to digital intermediation do show that growth, but those activities are relatively small. The large digitised products are obviously the import of database services and database originals and the import of goods via e-commerce as a sales channel. The import of database services and database originals does double in the 2010-2016 period, but the import of goods via e-commerce as a sales channel “only” grows by 38 percent. Because of the difficulty in finding the correct data to fit the scope of this research, the outcome should be regarded with caution. Changing the interpretation of the scope, can have large effects on the outcome. One of the avenues which should definitely be looked into in follow-up studies is the use of the sales of EU-based web shops as developed by Statistics Netherlands (CBS, 2020d) for the purchases of goods via e-commerce by Dutch consumers.
This study does not quantify the extra purchases or imports triggered by digitalisation. It only attempts to make transactions related to digitalisation more visible in existing statistics and to explore where data is available in existing statistics. This study is an explorative attempt to quantify a few of these relevant digitised activities. The intent behind the study is not to give a complete overview of all imports of all digitised products.
In conclusion, this study provides a first reasonable experimental overview of the size of a few digitised activities in the Dutch economy. The status of this study is experimental. Feedback on the study is therefore very welcome.
Further research will be needed to improve these estimates. This research could focus on possible improvements in survey design for better specification of digital activities. Alternatively, innovative methods such as use of web scraping and machine learning are very promising and could be further explored in the future.