Trust the Experts
Organizations should put competent people in charge of things and fire those who prove incompetent
Much of our government bureaucracy is structured around a deep-seated fear of someone somewhere doing the wrong thing. Whenever funding is proposed, it is almost always tied to a mountain of reporting requirements and strict controls out of worries of grift or improper usage. Whenever power is delegated, it is only after first carefully tying that power to a host of multilayered checks and veto-points. The motivation is obvious: people misusing funds and power is bad. The problem however is that this cure is often worse than the disease.
Talk to any leader, or even just a manager, at an organization with a track record of success and they can tell you that micromanaging isn’t just annoying, it also doesn’t work. This is not some new insight, there’s a Chinese proverb that, translated, goes something like “choose those you trust, trust those you chose (用人不疑，疑人不用)”. You can design institutions to try and prevent bad leaders from being part of the organization, and to provide mechanisms to kick them out if you decide they are not worth trusting any more. However, if you want people to be able to do things (which is presumably why the institution exists) you have to be willing to trust people enough to give them power and the freedom to exercise that power. If you build an institution in which incompetent people can be removed with relative ease, then people who have managed to rise through the ranks and distinguish themselves are very likely to be worthy of trust. If you keep people who nobody trusts around, and try to just restrict their authority so they can’t cause problems, you’ll end up with an institution composed of incompetent people that don’t mind doing nothing of value. If you’ve worked in a government bureaucracy, or in one of an (un)healthy percentage of large corporate environments, some of this may seem familiar.
So how do institutions end up like this, the government especially, when “don’t micromanage, just hire good people” seems like such common sense? There are at least two obvious candidates.
Loud vs silent costs
If a given department manager is caught embezzling $100M and is all over the news, that’s a loud failure. On the other hand, if 10,000 employees spend 10 hours a week at $20/hr (the average wage for a college graduate) filling out paperwork to prove that they’re definitely not embezzling money, that’s over $100M every year. However, that 1 person who gets caught embezzling will seem like a much larger deal than the dispersed and gradual wasting of funds due to the paperwork. As a result, organizations have built up huge loads of of work that could be grouped under the umbrella of “reporting requirements”. All this work is meant to reduce waste and corruption, and is often pushed by small-government conservatives, but it often has the opposite effect. It grows the size of the bureaucracy and very often is responsible for more waste than if the organization just accepted some risk of malpractice and responded by setting harsh punishments for those caught. Not only do these strict reporting regimes generate first-level waste in terms of labor hours, they also tend to scare off the sorts of competent people who actually want to accomplish something with their career. As the organization hemorrhages competent people, it becomes even more inefficient and wasteful. Altogether, reporting and compliance regimes have a variety of significant direct and indirect costs that are often underappreciated because those costs are low salience.
What does this actually mean for policy? Well, it means the optimal level of fraud is not in fact 0, because to achieve 0 fraud requires monitoring and compliance costs to approach infinity. When putting in place a compliance regime, we should actually sit down and try to determine how many man-hours of paperwork are being generated, what that will cost, and whether that’s greater than the risk of the thing the compliance is meant to mitigate. We also should be pragmatic in our evaluation of how successful our compliance regime will be in actually preventing fraud, given that the kind of person who commits fraud is also often the kind of person who will be willing and able to find loopholes in your compliance system.
Underrating the cost of inaction
This could in some sense be considered an extension of the above pattern, but with its own emphasis. When something new is done and causes damages, that is a very loud cost. For example, if a new freeway gets built and it pollutes all the houses around it. However, if the status quo is causing damages, that can be very quiet even when the damages are severe and the mitigating actions are straightforward. A classic example of this failure in institutional design is California’s CEQA. CEQA, which stands for the California Environmental Quality Act, is a piece of legislation designed to prevent projects that would hurt the environment of the area they are being built in. There are many things to complain about with regards to the design of CEQA, but the issue relevant here is how the law enshrines status-quo bias. CEQA only allows for stopping new projects (potential misuses of power), but not for proactively taking action to fix current problems. So if a 10 year old highway is found to be polluting a neighborhood and killing the residents, the government might propose replacing the highway with a train line. However, that train line must now go through review to show that it won’t cause harm, even though every day the train isn’t getting built is a day that harm is being caused.
Instead one could simply have a transportation department that is trusted with managing that infrastructure. Such a department would obviously have some limits on its authority, but those should be primarily used to prevent irreversible harms. For example the department probably should not be authorized to bulldoze housing without approval from representative political bodies. However, if the department decides to replace a freeway with a train line, or to build bike lanes across a city, they should be able to just go do that. If they screw up, leadership can be removed and the changes can be reversed. If that became necessary it would be very bad, but it would not be the end of the world. On the flipside, people who spend their whole life building a career would generally prefer not to be fired and disgraced, and so will avoid pursuing projects they expect to get them fired and disgraced. People inside of the organization have a strong incentive to work on projects that will bring prestige and popularity to the organization so that they might climb the ladder, since climbing the ladder means being given more authority. Instead what we often see is that the public sector hemorrhages talented individuals because anyone who is motivated will find themselves unable to make meaningful progress towards accomplishing the organization’s goals, and instead finds themselves endlessly arguing in front of review boards to try and get approval for proposed projects.
It’s important to reiterate that I’m not proposing no oversight. I am simply saying that we should carefully weigh the costs of oversight and restrictions on authority against the real world costs. We should focus on punishing actual harms vs preventing hypothetical harms, and we should view personnel selection as the first and most important tool in ensuring that an organization moves in a productive direction.
Conclusions
Our government is regularly failing to pursue stated policy goals effectively, and is wasting an rapidly growing quantity of time and energy on filling out paperwork in a vain attempt to further reduce the odds of fraud or malpractice. If we have more courage and are willing to tolerate even just marginally more tail risk, we can restructure our organizations to be vastly more effective and efficient. We can view the cultivation and empowerment of competent, trustworthy, and ethical people as a core part of our institutions and leave our children with a society of capable of rising to whatever challenges may come.
Alternatively, we can keep going like we are and hope that we finish the paperwork before our power grid fails and our planet is burning. At least the lawyers will be well paid.
