Theory X
This week, we’re continuing our exploration of the work of Sidney Dekker, particularly his book Drift Into Failure. Dekker applies complex systems thinking to issues of safety and organizational decline.
In our first essay in the series, we talked about why complex systems are that way. Complex systems’ behavior is the result of the interactions of many different components. The number of possible outcomes increases exponentially with each new piece added to the puzzle, which results in complete unknowability. Complexity simply does not compute. Dekker’s point is that the complexity of our systems - the size of our institutions, the amount of technology involved, etc. - has far outstripped our ways of thinking about them. The world is increasingly complex; our philosophy, not so much.
In our second essay, we discussed Dekker’s concept of drift into failure - the gradual, nearly-invisible and then all-of-a-sudden slide of systems and institutions into catastrophe. Dekker’s point is that systems naturally slide towards disaster for reasons which are entirely rational on the ground. Disaster is not necessarily the result of wrongdoing or incompetence; in fact, completely functional and efficient systems can, and often do, drift into failure. The nature of complex systems means that decisions which are locally rational (trade-offs against conflicting goals, non-ergodic low-risk temporary solutions, going along with well-entrenched and so-far successful company culture, etc.) can lead to outcomes which are globally disastrous.
These two ideas lead us to a necessary conclusion that represents one of Dekker’s most challenging ideas - and the one which, if accepted, stands to revolutionize how people work, manage, and communicate.
That conclusion is this:
When something goes wrong, it’s a waste of time to blame anyone.
In 1957, an MIT professor named Douglas McGregor gave a speech.
In that speech, he outlined the dominant theory of management. He called it “Theory X.”
Managers who believe Theory X believe that “employees are inherently lazy, not very bright, gullible, and self-centered. They believe employees generally lack ambition and resist change. They believe management’s job is to use rewards and punishments to direct, motivate, control and even modify employees’ behaviors in order to get organizational results.”
I have quoted here from the book Learn or Die, in which I first read about McGregor.
You can view Theory X as the “dumb ox” theory of management. Employees are the oxen; they are strong but slow-witted. It is the role of the manager to crack the whip and direct their movement.
It’s important to note that very few people claim to be Theory X managers. They see themselves as forward-thinking, collaborative, and “on the same team” as their employees.
What their behavior reveals is another matter altogether.
When something goes wrong, someone gets chewed out. When someone does something great, they get a strong pat on the back and their picture on the wall. When the numbers are down, the boss’s boss complains, and so the boss complains to his team and tells them to “get their asses in gear” and “get those numbers back up.”
The modern workplace, in other words, operates under a command and control model. Managers command in order to control the employees. Mistakes are punished, good deeds rewarded, so as to generate less of the former and more of the latter.
Before you judge these managers too harshly (and I’m sure you’ve had a few that fit this bill) - we are no different.
My kids are constantly leaving their shoes in the middle of the hallway. They dump their backpacks on the ground rather than hanging them up on the hook by the door. When they do this, it stresses my wife and I out, and we bring it up - “WHY can’t you guys put your stuff away?!”
When the CEO of a major corporation gets called to task in front of Capitol Hill, there is general rejoicing on the left. “Finally, these corporate fat cats get what they have coming!” After all, these people enrich themselves at our expense. They flout the spirit of the law by finding tiny loopholes and exploiting them to the utmost. Did you know that Tesla didn’t pay any tax last year?
Dekker rejects this thinking. All of it.
Yes, bad actors exist. But by and large, people are doing their best with the tools and knowledge they have. People want to do a good job. They want to be safe. They want to provide value to the people around them. People are good and mean well.
Complex systems thinking means that when bad things happen, the place to look for the cause is not the person who happened to be there when things exploded. After all, if you recall our essay from week two, everyone involved can make entirely rational decisions every step of the way and still end up in disaster. What’s more, decisions made years - even decades - before can influence the moment of collapse. In fact, when we think about the sheer number of possible system interactions that influence what happens in any given moment, the person who happened to be at the controls is the least likely to be the prime mover. What are the odds?
If the people involved are doing their best with the tools and information they have, and if systems interactions far away in time and space from the moment of disaster can play a critical role in producing that disaster, then it makes little sense to punish or reward the people on the ground. If they didn’t cause the event, it’s illogical to assume that you can either reinforce or inhibit future events by doing anything to the people involved. In fact, all the time you spend assigning blame and applying punishment is not just wasted but, even worse, prevents you from getting to the actual cause of the problem.
Dekker encourages us, rather than trying to find someone to blame - which makes us feel better but does nothing to make us any safer - to look at the systems involved. What information and tools did people have access to in the moment? How did those elements interact to make what happened rational, or even inevitable? What choices, often made far upstream, led to the environment being as it was in that moment?
This type of thinking is extremely challenging to implement for a few reasons. For one, it goes against what Rene Girard theorized was an in-built tendency to find a scapegoat - someone who can relieve the group of its responsibility for change by taking all the blame. Scapegoats, offered up in sacrifice, reassure us that not only is the problem solved, the problem was never with us in the first place. We are absolved, and life goes on.
The second reason this kind of thinking is challenging is that, followed through to its logical conclusions, the blame for negative outcomes often comes to rest with those in charge. Poor employee performance often gets tracked back to management decisions and company policies. Disasters often get tracked back to trade-offs against conflicting goals, unclear leadership, lack of resources or high-level policy. The waves that crash against our shore often stem from pebbles tossed into the water by the people we put in charge. Far easier for all concerned to point their fingers at the poor sap who happened to have their hands on the wheel in the moment.
People will often complain that Dekker’s philosophy excuses us of personal responsibility. I don’t know exactly what Dekker says to that charge, but I know what I would say: if he’s right, focusing on personal responsibility simply exacerbates the problem. The question is not what you wish were the case, but what is the case - and, if Dekker’s right, sticking our heads in the sand because we find things aesthetically distasteful is extremely unethical.
If we care about improving society, making our world safer, and combating the long-term problems that beset our planet and species, we need to understand the nature of the systems we’re working with - and how best to improve them. A moral sense of personal responsibility evolved to work in a world of small, genetically-homogenous tribes with limited technology may not be the best tool for navigating a vast diverse and interconnected world.
SOMETHING I'M READING:
An excellent primer on AI progression and the issues surrounding it. A great intro to the debate.