In PostgreSQL 18, table ua has columns (id, note) and table ub has columns (id, label). How many columns does SELECT * FROM ua JOIN ub USING (id); produce?

Explanation

USING suppresses the redundant join column, so the output is one id followed by note and label. The ON form SELECT * FROM ua JOIN ub ON ua.id = ub.id; keeps both and produces 4 columns, two of them named id - a difference that matters when a client indexes result columns by name.