Well someone pointed me to:
http://econlog.econlib.org/archives/2011/04/applied_ignoran.html

That’s worth spreading and has to be considered.