The resource availability hypothesis revisited: a meta‐analysis