Question: Which should I use ?PostgreSQL drivers for Unicode v ANSI ?   These options appear when I attempt to configure am ODBC  DSN  for a connection to a PostgreSQL database.

Answer: Some background information  as both ANSI and Unicode represent encoding formats - but when adding the PostgreSQL drivers both AND ANSI appear , so which driver to use? Use Unicode unless your application is not Unicode compliant.

It's worth clearing up some confusion about the ANSI name - which is actually an organization (American National Standards Institute) rather than an encoding format. But it is also used synonymously as a character set  aka Window-1252 strangely not recognized by American National Standards Institute. 

ANSI is used to encode Latin alphabet - in the one-byte format aka fixed-width format. Whereas Unicode (there are different flavours)  has a variable length format . Most operating systems are now fully compliant with Unicode - and versions of Operating Systems that were ANSI  have reached there end-of-life versions.    

ANSI is  maintained to deal with backward compatible applications. There are legacy applications not programmed  for Unicode .