Continuous Improvement Makes Everything Look Bad Looking Back

Here we are again in the spasms of self-flagellation about how we treat (or mistreat) people who have planned and sometimes carried out the murder of many of our citizens. We worry that revelations about harsh tactics used to get information from some of them may have damaged our international reputation and there are calls for a full scale investigation to uncover and reveal additional details. As long as we do such things – the argument goes – we cannot hold the moral high ground nor expect cooperation from others.  The ends don’t justify the means.  Actions speak louder than words. 

But actions must be framed and interpreted, and that requires words and analysis. Sometimes the reason something is done does make a difference and the some ends can justify some means. I believe we make a big error in framing our actions by demanding, and letting others demand, a measure of perfection not attainable among humans.   In those terrible times after 9/11, I think the U.S. showed amazing restraint, even after we captured some of those who planned the attacks that killed thousands of Americans.  Under passionate circumstances, and even under normal ones, mistakes are made.   Humans overreact, over respond and overreach, and things done in the passion of one situation may seem stupid or even evil after those circumstances have passed.  We go over and over our mistakes, often very publicly, and say that it is a sign of strength to do so.  We allow a successful program to be “ruined” by one mistake or even one insensitive action or even one remark that could be interpreted as insensitive. We may be acting honorably or we may be overlooking the fundamental nature of error and improvement. Maybe we are doing both.  

You have to refine and re-refine what you do to minimize the scope of errors and also – of you really want improvement – you have to minimize the finding of blame.  Even a very rigorous system cannot eliminate all error.  And we have an additional caveat. While this total quality approach is great for physical processes and assembly operations, it still doesn’t work as well in emotion or politically sodden human affairs and it especially doesn’t work when you have adversaries.  Focus on your errors gets to be like trying to understand a contact sport from only from one perspective.  Every contestant is going to make mistakes, get hurt and inflict pain.  If you fail to look at the whole picture, even the champion will look like the loser by those criteria taken in isolation.

We justifiably complain that we don’t live up to our own high standards.   But that is in the nature of complicated systems, especially human systems.  An after-action analysis will always find flaws.    Mistakes should be identified and corrected and then we need to move on, avoiding the twin errors of glossing over mistakes or being blinded by them.  Learning and improving only takes place in that middle ground between treating errors as terrible sins and ignoring them as inconsequential.

I want to be very careful to underscore that I am not advocating lowering our standards.   America should and does hold to the highest standards and we can only improve setting the bar higher than we can presently achieve.  But I think we open the door too far for criticism when we allow some of the nastiest despots and terrorist to assume the high ground of victimhood.  It is the old problem of moral equivalence.  A man who takes a pencil from his office and the one who embezzles a million dollars are both stealing from their employers. But they are not really the same.

Every judgment needs to include the “compared to what?” question. If we allow the frame to be a comparison to some theoretical perfection, we will always come up short. We can always imagine something better than we can achieve.  And ironically the more we work to improve – i.e. the higher the standards we set – the worse we look in relation to our own every rising goals.  The more positive achievement you make, the worse everything else looks.