11. February 2025
After the holiday season, some people return gifts which they have received.[1] Using electronic payment data, we want to find out how pronounced the refunds are across different expenditure categories and how large the differences are in January (when more returned items are likely to have been gifts) relative to the rest of the year. Furthermore, we are interested in whether the size of refunds has changed over time and whether the transaction channel plays any role.
To estimate the size of refunds, we use the expenditure data provided by Worldline Schweiz AG, which are used for the dashboards of Monitoring Consumption Switzerland (MCS). We focus on the merchant categories Clothing, Department Stores, and Electronics.[2]
Figure 1: Evolution of refunds in percent of the sales volume for the categories clothing, department stores, and electronics together, both in ecommerce (ECOM) and point-of-sale (POS). The line shows the 28-day moving average.
Figure 1 shows the evolution of the 28-day moving average of the refund share (refunds in percent of the sales volume) for these three merchant categories together since November 2021. We make three observations. First, refunds peak in January. Second, refunds amount to approximately two percent of the sales volume on average in the Worldline data. Third, refunds in percent of the sales volume have decreased over time until 2024 before rebounding somewhat in 2025. More detailed inspection of the data reveals that the downward trend was driven by a decrease in refunds whereas the sales volume remained quite stable over time.
Figure 2 sheds further light on the increase of refunds in January by distinguishing the three expenditure categories, considering both point-of-sale (POS) and ecommerce (ECOM) transactions. We find that refunds in department stores drive most of the increase of refunds in January. The refund share is around 0.8 percentage point (pp) higher in January than in December. For electronics and clothing stores, there is also a slight increase of refunds in January by 0.1-0.2 pp.
Figure 2: Evolution of refunds in percent of the sales volume around the holiday season across categories (ECOM and POS). The figure displays the mean of the 28-day moving average since November 2021.
Figure 3 investigates the differences in refunds across transaction channels per expenditure category. When considering point-of-sale (POS) transactions, we find that refund shares are small and very similar across categories. For ecommerce (ECOM) transactions, however, refund shares are much larger and more heterogeneous across expenditure categories. ECOM transactions for clothing have a share of refunds that is over ten percent, with peaks as high as 16 percent. Refunds in ECOM are lower for the other expenditure categories but still considerably larger than for POS transactions. Because ECOM transactions account for a relatively small fraction of transactions, however, Figure 3 shows that the refund share for both ECOM and POS transactions together is quite similar across expenditure categories, as is the case when we only consider POS transactions.
Figure 3: Refunds in percent of the sales volume across categories since 2023, by payment channel. The bars show the minimum and maximum of the 28-day moving average of refunds in percent of the sales volume.
These findings suggest that shifts of transactions from POS to ECOM may be associated with higher costs because of more refunds, both for firms and for society. Some of the costs associated with parcel delivery, such as pollution for example, may neither be fully internalized by firms when they price the delivery nor by those who make presents. Of course, such costs must be compared with the carbon footprint associated with store visits and POS transactions.
Should we worry about the positive effect on the refund share for the considered categories due to the shift from POS to ECOM transactions in recent years? Figure 1 reveals that other forces have outweighed this effect, generating smaller refund shares in 2023 and 2024, also at the peak of refunds in January. Indeed, the January gift hangover seems to have become smaller compared to 2022.
[1] See for example Retouren Resale Rückgaben – Galaxus and Nach Weihnachten sind besonders viele Retouren unterwegs | tagesschau.de.
[2] Clothing includes the merchant category codes (MCC) 5137, 5139, 5611, 5621, 5631, 5641, 5651, 5655, 5661, 5691, 5697, 5698, 5699. The MCC 5311 defines department stores and electronics cover the MCC 5045, 5732, 5734, 5946, 5997.