Efficiency vs. Robustness

Efficiency and robustness are enemies. The more you have of the one, the less you have of the other.


Efficiency means that every part is producing an effect. Robustness means that damage can happen and the effect can still be produced. The only way that it is possible to take damage and keep productivity up is if there was some spare capacity which was activated to take the place of what was damaged. That spare capacity was, by definition, inefficiency.

(The subject is sometimes confused by there being standards of robustness, and efficiency being contextually defined to be having no more spare capacity than that standard; this context-specific usage is not what I’m talking about. I’m talking about efficiency and robustness as such.)

You can see this trade-off very well in birds. Birds, in order to fly, must be extremely efficient. The price that they pay is that they are very fragile. Their bones are famous for being hollow, which is awesome, but their bones are also extremely easy to break with blunt-force impact. Birds are very muscular with little fat, but the cost of that is that they need to eat frequently and can starve to death quickly.

This same thing applies to large organizations. Young people are often astonished at the inefficiency of large organizations. They are right that large organizations tend to inefficient. What they don’t realize—because they’ve never experienced it—is how much inefficiency is required for the organization to survive when things go bad. And things can go very bad.

Probably the most frustrating example of this is the small amount of trust typically placed in employees and in consequence the difficulty they have actually getting work done. This must be understood in the context of how many problems a single employee can create.

It is not merely a rogue employee who intends to do evil which is the problem. They can cause truly enormous amounts of damage, but they are also vanishingly rare. Far more common are things like:

  • Taking vacation
  • Getting sick
  • Personal problems
  • Leaving for another job
  • Incompetence

All of these things create problems, and some of them are so predictable as to be reliable (like taking vacation and getting sick). The result is that organizations which have protections against these things tend to be far more robust and thus last longer.

These protections will tend to look like:

  • Having multiple people who can do the same job and are sufficiently involved with the particular job that they can take over while the other person is away or otherwise not doing his job.
  • Requiring non-trivial decisions to be filtered through at least one other employee, so that there is a sanity-check in place.
  • Requiring extensive documentation of what’s going on so that others can take over
  • Using tools which, while not the best suited to the task, are wildly used, so that replacing a person who uses them is easier
  • Having a corporate culture, created by things like too many meetings, so that people tend to solve similar problems in similar ways, so that people are more replaceable

All of these things are inefficient, but they do create robustness.

It should be noted that there is something of a catch-22 here, too. These practices are not pleasant to work under, so they tend to create a culture of zero loyalty to the organization, and in many cases encourage people to leave to find something better elsewhere. In more extreme cases, they actually create a selective pressure for only awful people who couldn’t get hired elsewhere to remain, which is a vicious cycle since they make the workplace worse for competent people.

I want to be clear that I am not saying that these things are actually good, or a good way to run a business. I’m only saying that we should understand that they do exist for a reason, and it is important to know what that reason is if one wants to try to prevent smaller companies that are growing from falling into the same traps.

I also want to note that not all behaviors of large companies are ascribable to this cause. In some cases, I think that there are behaviors which are analogous to cancers and autoimmune disorders.

An example of the former would be divisions which pursue growth over profitability and suck resources away from profitable divisions, often to cause enormous damage when they fail and have to downsize.

An example of the latter would be an accounting department becoming so hyper-active in combating waste that they make it nearly impossible for employees on trips to get reimbursements for anything, resulting in lost opportunities and costly workarounds.

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.