I think the argument is: If it costs $4 to improve pensions $1 than let's find out the best way to spend the $4. Maybe an option that only costs the company $3 but nets us $2 is worth listening to.