The one thing that separates the good from the bad is their ability to test their own work.

A programmer who thinks about everything that they will need to test for covers all of the possible outcomes and sends out code that is cost effective.

A programmer who focuses only on the primary goal does not cover all eventualities and turns out code that has holes in it. This is expensive. If code with flaws get out to the testers it is far more expensive to fix.

Of course the latter can eventuate when management concentrates on cost instead of quality and of course this ends up being far less cost effective.