I tested more since yesterday and I realized that the reason I see different results in SAS 9.4 and EG is due to the fact that the session encoding in EG was set to wlatin1 instead of UTF-8.   When I set the session encoding to wlatin1 in SAS 9.4 or EG I get better data quality than with utf-8, even though the Postgres shows this: Server Encoding: UTF-8, Client Encoding: UNICODE    When I use SAS in wlatin1 characters like bullet points • and É are displayed correctly. When I use SAS in UFT-8 I get in both cases this: �   I don't understand that.    ... View more