Lam’s book Fewer Rules, Better People: The Case for Discretion is filled with specific examples of cases where the absence of discretion compels people to take actions contrary to justice or even just common sense. In one chapter, he tells of a woman who, through her organization, was hosting a roundtable conversation at a conference. She wanted to order coffee for the event, but her institution had a rule that all catering orders must go through a particular vendor. But there was a problem – the roundtable began at 9:30, but the vendor didn’t even start taking orders until 10:00. However, there was a Starbucks right around the corner, and it turned out that the vendor in question subcontracted out coffee orders to Starbucks – indeed, to that particular Starbucks location. So the woman hosting the event tried to get her institution to just let her place the order directly with Starbucks, to no avail:

Surely, my host argued, this was enough evidence that this coffee purchase was within the spirit of the rules. The administrator disagreed and did not approve the purchase. It was against the rules.

It might be easy to dismiss this sort of rule-mongering as being little more than annoyance, but not indicating anything substantial. But Lam disagrees. He thinks that being ruled over by administrators dedicated to unyielding submission to policy manuals and rulebooks can be as corrosive to the fabric of civil society and moral development as life under tyranny:

Tyranny has done as much as anything else to keep humans from flourishing. Western liberals have been less concerned with a figure who is the opposite of the tyrant in disposition but no less to be feared. This is the by-the-book bureaucrat. These bureaucrats are naturally inclined toward legalism and deeply afraid of exercising discretionary judgment. They do not want the responsibility; they fear all blowback so they are highly risk averse. Confronted with a decision that they are unsure about, they look up the governance language and are comforted when they find that the issue is out of their hands. If someone goes to the by-the-book bureaucrat with a novel idea, the only way to get them to yes is if there are explicit rules in favor of the idea.

A critical capacity for us as individual people, and for developing a flourishing society, is the ability not merely to recognize a rule and know how to comply with it – it is the ability to understand the reason and purpose behind the rules. Rules are not self-justifying simply because they are rules – by their very nature, rules are meant to exist in service of some other, larger purpose. A commitment to live one’s life by simply applying the rules, whatever they may be, blunts our moral development and our capacity to exercise virtue. This isn’t just true of the by-the-book bureaucrat. It’s also true for the citizens whose behavior is determined by subservience to legalistic algorithms:

But I think that even worse than this is the effect such societies have on the obedient. The goal of surveillance-state legalism is to turn all citizens into compliance robots and every bureaucrat into a by-the-book bureaucrat. It does this by turning human motivation in all of its rich complexity into fear of stepping out of line and into a love of acting for the sake of rules. This is the true horror of legalism.

He makes a personal analogy by reflecting on his own responsibility as a parent to teach his daughter not merely to follow a list of prescribed actions, but to understand the why behind those actions:

The aim of creating responsibilities and holding her to them is not for her to act for the sake of the rules and the punishment and rewards they bring. We want her to feed, water, and brush her bunnies because they are vulnerable living things whose well-being depends completely on their caretakers. We want her to empty the dishwasher in a timely way because it is important for a kitchen to be clean and useful, because it is essential not to be a free rider in the home, and because you have obligations to other people. We want to cultivate in her the spirit and reasoning behind the rules, especially when they are good ones.

We also want her to have the judgment to figure out when rules are not good, either because the rationale behind them is flawed or because she has figured out a better way to do things.

He admits that it might be simpler to simply legalistically enforce the rules on his child. But that would be a failing on his part:

Make the penalties severe and consistent enough and I’m sure I could get better, more consistent compliance. But if my child ends up being the administrator who refuses to approve the coffee purchase, I have failed to raise a decent, reasonable person as much as I would have failed if I had raised a tyrant. Choose between a world of apparatchiks who follow the letter of the law and a world of imperfect decision-makers who have judgment and motivation to do well by the spirit of the rules, who are ready to identify better ways to do things, I’ll take the latter 100 percent of the time.

Not only does legalism diminish our ability to develop as moral agents, it also leaves people unprepared for, and incapable of, acting in the face of situations the rules didn’t or couldn’t account for:

Novel circumstances will require judgments based on the reasons behind the rules. People who live out of fear of noncompliance will have no idea what to do in such circumstances. Even worse, they will apply rules in ways contrary to the reasons those rules exist in the first place. They will act unjustly under the cover of law, whether it is in denying coffee or denying freedom.

Lam is also particularly concerned about the growing tendency to make decisions via AI generated algorithms as a substitute for human judgment. Use of AI algorithms in decision-making robs us of even the possibility of understanding the reason behind the rules:

These are equations for which there is no explanation in human language. A programmer can show you this equation, though you could never read it for its length, and the only coherent description of what it means is “this is the equation that fits all the past verdicts the best.” Deep learning is not just opaque, it is humanly indecipherable. Deep-learning rules are the logical endpoint of the proliferation of rules. It is the endpoint of legalism, where the laws of bureaudynamics take us in the most extreme case, where no one even understands the rules that govern us.

Thus the use of AI generated rules does little to alleviate the problems of the by-the-book bureaucrat:

AI is like any by-the-book bureaucrat who, settling on the rule that prohibits coffee purchases from any vendor but A, will give you no explanation as to why vendor A can purchase Starbucks coffee but you cannot. Asked why that rule is the right rule, they only insist that it is the rule.

Ultimately, Lam believes that Han Fei reverses the true relationship between mediocrity and rules:

Han Fei believe that legalism was the antidote to mediocrity. I think legalism is the cause of it. Standardized food, standardized homes, standardized essay grading, and standardized bureaucrats are at best okay, merely passable objects to be tolerated. They are never excellent, and they do not inspire excellence.

But this discussion calls for more than just second-order, abstract discussions about why discretion needs to gain ground against legalism and rules. For the ideas in the book to be meaningful in a practical sense, the rubber will have to hit the road at some point. To that end, Lam has a series of ideas about how and where to expand the role for discretion. I’ll cover that in the next post.