While writing Some SQLite things I recently learned a while ago I bumped into SQLite extending the SQL syntax allowing SELECT column aliases in WHERE and ON conditions of JOIN clauses whereas standard SQL only allows it in the HAVING clause (some other DMBS allow this as well).
It reminded I never wrote a blog post about WHERE versus HAVING, which can be phrased into just two points:
WHEREis applied before theSELECTis assembledHAVINGis applied after theSELECTis assembled
This also means WHERE has a lot more influence on performance than HAVING, so using HAVING just so you can use aliases there instead of in the WHERE clauses is not a smart idea unless you verify in the query plan and query optimisation output of our DBMS this has no influence (some DBMS even allow you to see the results of query rewriting which you can use to verify if two queries really are equivalent from the DBMS perspective).
Since HAVING can be used without GROUP BY, many people work around the WHERE alias limitation. Given the above effects on performance, this is not a wise idea, despite the same end-results. See the example below that uses the [Wayback/Archive] MySQL :: Sakila Sample Database: customer table.
Luckily there has been enough other material explaining WHERE versus HAVING, so I can reference it from here, for instance (note note limited to MySQL/MariaDB) [Wayback/Archive] mysql – WHERE vs HAVING – Stack Overflow (thanks [Wayback/Archive] baloo and [Wayback/Archive] Fishdrowned):
The Wiert Corner - irregular stream of stuff 