Inconsistent data type behavior for date field in dataset

Dear Support Team,

we have received the following request from a user regarding dataset https://data.bs.ch/explore/dataset/100120/, which we would like to forward to you for further investigation.

The user reports an inconsistency in the handling of the field vu_jahr:

• The field is defined as a date in the dataset schema.

• However, when exporting the dataset, the field appears to be delivered as a string.

• When querying the dataset via ODSQL (using a SELECT vu_jahr GROUP BY vu_jahr query), the result returns proper datetime values (e.g. 2011-01-01 00:00:00+00:00, etc.).

This inconsistent behavior causes issues in the user’s synchronization process. While they have implemented a workaround on their side, they asked whether this could be fixed at the platform level.

Example query used by the user:
https://data.bs.ch/api/explore/v2.1/catalog/datasets/100120/exports/csv?lang=de&timezone=Europe%2FBerlin&use_labels=false&delimiter=%3B&select=vu_jahr&group_by=vu_jahr

Could you please check whether this behavior is expected and if there is a way to ensure consistent typing of this field across exports and API access?

Many thanks in advance for your support.

Kind regards,
Boris