Chesterton's Fence

Or, When to Step Back and Think About Complex Problems

Imagine that you’re working on a project in the middle of the country. You stumble upon a fence on a plot of land. The fence doesn’t immediately appear to have any real purpose or sense of being there — it’s just getting in your way. Should you tear it down?

No, says GK Chesterton in his book The Thing. In fact, he says if you don’t know why it’s there, all the more reason to leave it up and work around it until you have a better sense of why exactly the fence is there:

In the matter of reforming things, as distinct from deforming them, there is one plain and simple principle; a principle which will probably be called a paradox. There exists in such a case a certain institution or law; let us say, for the sake of simplicity, a fence or gate erected across a road. The more modern type of reformer goes gaily up to it and says, “I don’t see the use of this; let us clear it away.” To which the more intelligent type of reformer will do well to answer: “If you don’t see the use of it, I certainly won’t let you clear it away. Go away and think. Then, when you can come back and tell me that you do see the use of it, I may allow you to destroy it.

This idea of reform vs. destruction is more commonly known as Chesterton’s Fence and I like to think of it as a mental model to go alongside the Nirvana Fallacy. Where knowledge of the nirvana fallacy keeps us from falling into a status quo bias, Chesterton’s Fence implores a bit of humility and asks us to take a step back and think in terms of “reform” rather than “replacement,” at least until we hit a certain epistemic threshold.

This is a particularly helpful foil when thinking about solving very longstanding and complex problems, especially those that have some kind of cultural purpose. Solutions rarely come about as a function of top-hatted billionaires sitting in smoke-filled rooms and plotting some kind of conspiracy to create a terrible policy or product or system that will hurt lots of people. Rather, they’re spontaneous orders, the products of human action but only of human design on the smallest components. At the most meta-level of norms and mores, these things look more like language than they do formal rules developed in an academy. Hayek’s “Cosmos and Taxis,” a decidedly not conservative document, is one of the first to make me take a step back and realize that most of the saber-rattling that radical replace-not-reform movements have can have terrible consequences.

Meta Examples: Culture is Complex

Classic meta-examples of not following the warnings of Chesterton’s Fence are the horrors of the French Revolution and the fruits of new secularism today. The French Revolution was so radical in tearing down these fences that the Jacobins went as far as creating an entirely new calendar. The systems and norms they tore down only laid the groundwork for the Reign of Terror and eventually the rise of Napoleon.

I would go as far as to say that we’re living through a modern-day fence-destruction in how quickly Western societies have made useless religious institutions. In part as a rejection of supernaturalism and in part as a rejection of the conservative norms that these institutions brought with them (although I don’t think the sexual revolution explains everything here), the average American has decided that he no longer really needs to go to church, no longer needs to confess his sins, and no longer needs to teach his children about basic tenets of Western civilization encapsulated in the Judeo-Christian texts. It’s hard to draw a causal relationship, but this is likely a factor in increasing atomization, decreased optimism, decreased life satisfaction, and a whole host of other issues.

Practical Examples: Politics, Education, Technology

A few quick practical examples would include areas of politics, education, and technology generally.

Within politics, you’ll often see a push to clean up or destroy political machines. I see this most often from friends and acquaintances who are sick of losing with respect to their preferred political outcomes and decide to get up and leave to somewhere better. While exit here is likely preferable to destruction, there’s a complex game being played inside the political machines and just getting up and leaving doesn't 1) guarantee you won’t run into a political machine elsewhere; and 2) solve the initial problem of the political machines.

Political machines are, in my mind, complex regulatory mechanisms that allow a city or region to run with the interests of the people who are most invested in that region in mind. Sometimes this is good, sometimes it’s bad. Political machines are merely tools, not something good or bad in and of themselves. But what happens when you decide to clear out the political machine and just have “democracy” or whatever the purest equivalent is? I’m not convinced it isn’t either complete chaos, with the people who were formerly most invested in the political machine now having to resort to either outright corruption or grey markets, or something far worse: credentialist “meritocracy.”

Ask yourself: would you rather your city be run by the guy with deep local connections who is elected due to these connections (some of which you don’t particularly like) and his local knowledge or would you rather your city be run by a group of Georgetown MPPs and their JD friends who have been grooming themselves since 10 to run for office? Which do you think would actually better serve the interests of the people in the city?

Another example, one less fraught with peril and perhaps more applicable to a reader’s life, would be in education. I am one of the first people you’ll find saying that education needs to be reformed — I go as far as saying that school in and of itself has adverse effects on development and ambition. But it’s a mistake to come to frustrations in the educational system and imagine that that must mean that the whole system is silly or stupid or must be dismantled.

I actually think this is well illustrated by the reformer-destroyers of yesteryear when it comes to education: the folks who went to Prussia, studied the Prussian industrial education model, and brought it back to the States. It would have been easy for them to look at things like how much time children spent at home working around the farmstead or the market as inefficiencies and as time that could have been spent in the classroom. But this time away from the classroom is what made it possible for the family to run family businesses or a family farm. It’s no wonder that within a few generations — around the time of the Boomers — the desire or ability to take over the family farm or family business was essentially eradicated.

Today, I worry there’s a similar push with the desire for universal Pre-K or in the cultural practice of both parents working and having their child watched by strangers at an “educational” daycare center. Children, especially very young children, learn a lot from unstructured interactions with other people.¹ They need this time to play and to have totally unstructured activity time with other young children. This is how they learn social rules, norms, and mores that are otherwise difficult or impossible to make teachable in the classroom. This time that otherwise looks “wasted” is actually a Chestertonian fence that shouldn’t be torn down, as we don’t know what the alternative outcome will look like.

And that brings me to technology generally. I consider myself a techno-optimist. I want to see a future where people live for 100+ comfortable years, where we feel that scarcity is a thing of the past, and where we think about taking our annual vacations to Mars. But I also don’t want smart and talented founders and prospective founders wasting their time on solutions in search of a problem. The over-academicization of everything is partially responsible for this. So much technology has to come out of university tech transfer programs that millions upon millions is spent on solutions looking for problems.

More simply, it’s not uncommon that I hear a pitch for a problem that I hear a lot. But nobody has been able to come up with a solution to that problem that seems to stick.

This tends to mean there are one of two scenarios going on.

Either:

1) This problem is so incredibly hard to solve that the person who does solve it will almost certainly become a billionaire; or,

2) This is a pseudo-problem — it looks like a problem but there’s actually a reason it exists.

The best way (that I’ve identified) for a founder to differentiate between these two is to get as far down into the weeds as possible into what the problem really is. Is there a very high-resolution way you can test for that problem and get people to part with some money or time to try your solution?

I’ll probably go deeper into this with a post on revealed vs. stated preferences and how this interacts with that foil, but a key takeaway here is that sometimes things that look like problems really aren’t. They may simply be imperfect solutions for even bigger or more unpleasant problems and trying to solve for them may be a fools’ errand.

(I see this a lot with marketplaces, especially those trying to disrupt broker-client relationships. Brokers often exist because they have a set of complex tacit knowledge that would be difficult to institutionalize in a software product.)

The key point of knowing Chesterton’s Fence is simply to take a step back and examine why frustrations or apparent problems may exist before launching into finding solutions. Get as granular as you can about what the problem is and try to ask yourself, “why might this annoying thing exist? Is it trying to solve something that I don’t understand myself?” And go from there.

1. See Peter Gray’s excellent book Free to Learn for a lot of information on this.