How to stop social dysfunction: wide vs. narrow problems

Many social problems can be sorted into two groups—wide problems and narrow problems. Treating one kind as the other creates catastrophe, but we do it all the time, from policing to politics.

Thank you for reading The Garden of Forking Paths. This edition is for everyone, but if you’d like to support my research and writing, please consider upgrading to a paid subscription for just $4/month, or check out a copy of my new book, FLUKE.

If you’re in the UK, I’ll be speaking at the Winchester Books Festival in April.

---

Most social problems can be divided into two types: wide problems and narrow problems.

You’ve probably never heard of these terms. That’s because I just invented them.¹

But once you recognize the distinction between these two kinds of problems, you’ll better understand all sorts of social systems, from politics and the police to Ivy League admissions, cancer research, and bomb defusal specialists.

Many social ills are caused by wide problems being treated as though they’re narrow problems. When that happens, we have mistakenly engineered an avoidable disaster. And we do it all the time.

So, what’s the difference between a wide problem and a narrow one?

Selection pressures and self-selection bias

To understand the distinction, you first need to know what a selection pressure means. This term, which has its roots in evolutionary theory, refers to a system in which a specific trait is “selected for,” because it helps an organism solve a problem, or even to survive an otherwise deadly environment.

Imagine you’re a bird and you live on an island with most of the available food hidden away in slim, rocky crevices. If you can get to the food, you’ll live, but if you can’t, you’ll die of starvation. In that environment, there’s a strong selection pressure for long, narrow beaks that can fit down into the tiniest cracks, extracting the maximum possible food. That same pressure cuts against those with broad, short beaks, because birds with those traits will, on average, die off.

Human systems have selection pressures, but we also have varying levels of self-selection bias. This is where we choose to opt-in, or opt-out, of various social systems based on our traits.

Take a high school basketball tryout. If you’re the shortest kid in the class, there’s nothing stopping you from attending the tryout, but on average, those students are less likely to try to make the team. By contrast, the tallest kid in the class is more likely, on average, to decide to give it a shot. As a result, there’s a strong selection pressure for high school basketball tryouts on a few traits: height, athleticism, hand-eye coordination, an affinity for sports, and so on. Students who show up for the tryout that don’t possess those traits are unlikely to make the cut.

To put this in more formal terms, that means that high school basketball tryouts have both self-selection for who shows up and then strong selection pressure in terms of who actually makes the team. The group of students who try to make the team are extremely skewed relative to the entire student body. Then, within that already skewed group, the coach is only selecting those who are tallest, most athletic, and so on, to determine the team that ends up on the court.

But not all courts are the same. Randomly selected juries in criminal courts are designed to be the polar opposite kind of social system from a basketball team. The entire point of a “jury of your peers” is that there isn’t self-selection or strong selection pressures when it comes to who ends up as the twelve people deciding your fate. They’re supposed to be average people, by design, not unusual ones.  

The random selection into the potential jury pool is therefore supposed to avoid any skews from the overall population. There are much weaker selection pressures than a high school basketball tryout.

These two extremes give us useful archetypes for understanding wide vs. narrow problems.

Wide problems require:

• Weak self-selection

• Weak selection pressures

• As a result, the group ends up being a broadly representative sample of the population

Narrow problems require:

• Strong self-selection

• Strong selection pressures

• As a result, the group ends up being extremely skewed relative to the rest of the population

If you create a mismatch—solving a narrow problem as though it’s a wide problem, for example—you’ll be in trouble. Imagine buying the Los Angeles Lakers basketball team and selecting your team randomly from the population to make the squad representative of the city. It wouldn’t go well on the court.

Now, that sounds absurd! Such idiocy would never happen! Well, I’m afraid I’ve got bad news for you, because many of our most important social systems have a mismatch, in which a wide problem is being treated as though it’s a narrow one.

Politics and the police: wide problems, narrow solutions

Consider the worlds of politics and the police, which are responsible for significant abuses of power from bad apples that harm society. Both realms are, generally speaking, best treated as wide problems.

In other words, it would be good for society if politics was a big tent, in which people from all walks of life—and a relatively normal set of traits—were those who made decisions about how we should be governed. Sure, you might want some weak selection pressures to ensure that politicians were more intelligent, educated, and compassionate than the general population. But you wouldn’t want the group to be too skewed, especially toward undesirable, anti-social traits.

What have we done instead? We’ve engineered social systems that have rabid self-selection and selection pressures on steroids for the worst possible traits to attract and promote all the wrong kinds of politicians.

Which traits make it more likely that you’ll run for office? Consider the fact that the phrase “power-hungry” functions as a character insult, but what it literally means is someone who disproportionately covets power. Those are, by definition, the people most likely to seek political office.

Narcissists gravitate to the political spotlight, too. Psychopaths have a disproportionate need to control others. All these traits are part of the self-selection of who tries to become powerful in modern society. This is why there are so many weirdos and freaks and cartoon villains who enter politics.

As I previously noted:

Consider people like Matt Gaetz (R-FL), with his big vampiric hair, coiffed with one bottle of gel per outing into the sunlight, or Steve Bannon, who clearly thinks showers are a deep state plot, or Lauren Boebert, who constantly laments that Colorado law doesn’t allow her to marry an AR-15.

These are not normal citizens.

In Britain, the odds that someone like Jacob Rees-Mogg—a bizarre figure complete with top hat accompanying his 19th century attire and social views—would end up on a jury pool are basically zero, because he’s one of the most unusual people in the country. We get weirdos and cartoon villains because our political systems attract and promote them, amplifying a freakishly narrow slice of the population when a wider slice would be much better for everyone.

We then compound the problem because we treat candidate selection as a narrow problem, when it should be handled as a wide problem.

All of us want candidates who represent a broad slice of the electorate. What do we do instead? We design candidate selection systems that rely on an extremely unusual, extremist slice of the population to choose who will end up on the ballot through primaries, caucuses, or internal party member votes.

In most American primaries and caucuses, turnout is between 10 and 30 percent of the population. Those who vote are disproportionately interested in politics, often more ideological than average, and hold extreme views. In Britain, Liz Truss only emerged as the prime minister (for 49 days, RIP, lettuce pray) because she needed to win a tiny, unimaginably unrepresentative slice of the population: she got 80,000 votes from paid up Tory members, in a country of 65+ million people.

Once someone ends up on the ballot, they need to win the election—and try to cling to power for as long as possible. That selection pressure rewards other undesirable traits: those who are Machiavellian and are disproportionately ruthless and self-interested, are more likely to get elected and survive in politics.

The same is true with policing, a realm that would be much more effective if more beat cops were reflective of the communities they police. For example, in Ferguson, Missouri—one of the flashpoints in America’s racialized police abuse—two of three residents were Black, but eight out of ten cops were white. But even aside from race, those cops weren’t representative of the population, either.

As I wrote in Corruptible:

“If you’re a bully, a bigot, or a sexual predator, policing is a really attractive career choice,” says Helen King, who served as assistant commissioner of the Metropolitan Police in London. She’s right. There’s considerable evidence, for example, that domestic abuse by police officers is a significant problem. Some argue that such abuse is correlated with an intense, high-stress job. But other intense, high-stress jobs don’t seem to have similar levels of domestic abuse. There’s a more convincing explanation. Perhaps some abusive people are drawn to a powerful occupation, such as being a police officer, where it’s easier to get away with abusing others.

Many admirable people are drawn to serve in the police. But the badge and the gun functions as a disproportionate magnet to abusive, power-hungry people, creating strong self-selection—and the misogynistic, macho cultures that plague many corrupt police departments create correspondingly strong selection pressures that purge those who aren’t willing to conform to that culture. In the most dysfunctional police cultures, the good cops either become bad over time, or they end up getting culled, like a broad-beaked bird on a narrow-beaked island.

That creates a vicious cycle, as normal people don’t choose to self-select into heavily skewed, broken systems. They just do something else.

Policing and politics are therefore both wide problems. Sadly, we’ve designed systems that treat them as though they’re narrow problems. That mismatch often produces catastrophic results.

Narrow problems should have narrow solutions

However, many realms of our social world are narrow problems, not wide ones. These differ from politics and policing, because the best solution actually requires a skewed population, not a representative one. And it’s not just basketball teams, either.

Take cancer research, for example. I don’t want cancer scientists to be average joes. I want them to be weird in a variety of ways: hyper-intelligent, passionate workaholics, geniuses almost demonically possessed to find a solution to this terrible scourge that kills so many of us—nerds who can only be satiated when they have a medical breakthrough. If the trait pool of cancer researchers in your lab looks like a bell curve of the population, you’ve screwed up badly. It’s a narrow problem that requires a narrow solution.

Some realms are hybrids, in which different roles in the same system require varying degrees of narrowness. It would probably be better for humanity if the larger pool of soldiers in national militaries were more representative of their overall populations, but I don’t want bomb defusal specialists to have a middling, average ability to operate during a stressful situation. They need to be freaks on that trait, but the good kind that, when harnessed, can help avoid explosions.

Similarly, the world of news and newsletters flits between wide problems and narrow ones, depending on context.

You—the readers of The Garden of Forking Paths—are highly unusual people relative to the general population. I don’t know all 20,000+ of you, but I’d wager that the group of you is disproportionately curious, better educated, more English speaking, and that you read more books than the average person in the general population (you’re welcome for buttering you up).

But that’s a blessing rather than a problem, because the function of this social group is to provide brain food to intellectually curious people about an array of topics that don’t always interest the entire population. The narrow skew is good.

By contrast, information pipelines about our world in the form of general news should probably be mostly treated as wide problems. When news production becomes heavily skewed by partisan demographics—as with Fox News, for example—then people get sorted into narrow groups, which causes news to chase those viewers, which further radicalizes the narrow group, in a vicious cycle.

Similarly, when it comes to paid newsletters that are exclusively about politics, there’s a risk of audience capture—in which the writer feels the need to pander to that unusually partisan group, because otherwise they’ll lose money.

In other words, treating general news production as a narrow problem can warp our understanding of the world in dangerous ways, whereas niche information pipelines about specialized interests (environmental newsletters, for example, or industry news about advances in artificial intelligence for people fascinated by machine learning) are no problem at all.²

The internet has therefore produced a wonderful, but sometimes dangerous innovation, allowing people with niche interests to find each other, which is great when it’s used to apply to narrow problems—but disastrous when the same dynamics are applied to wide problems.

Sometimes, narrow problems are rightly solved with narrow solutions, but everything goes wrong because the strategy uses the wrong selection pressures.

For example, deciding who to admit to Ivy League universities is a narrow problem. It makes sense that Harvard doesn’t just take a random sample of the US population that’s 18 years old. Most of those students wouldn’t perform well and would drop out.

But the narrow solution that would work best—selection pressures around intelligence and dedication to academic work—has been partly hijacked by the wrong narrow selection pressures, such as being a legacy candidate (having a family member who previously attended Harvard) and socioeconomic class (because students from more affluent backgrounds are more likely to end up at Harvard).

The problem with Harvard admissions, then, isn’t that a wide problem is being treated as though it’s a narrow one, but rather that some of the selection pressures are sorting students into the incoming class using the wrong skew.

Matching problem and solution types to avoid Liz Truss

As I’ve repeatedly highlighted, just about every group of humans is a non-random subset of the population that has been skewed in some way.

The key is figuring out whether the issue you’re trying to deal with is a wide problem or a narrow problem—and then determining which selection pressures are sorting people into that system, for better or worse (height on a basketball team, power-hungry narcissism in politics, and so on).

The best social systems correctly match the problem type with the solution, using tailored selection pressures closely aligned to the goals within that system.

But when a mismatch takes place, well, that’s how you get a world run by people like Steve Bannon and Liz Truss. And I, for one, would prefer if we didn’t have to live in that avoidable dystopia.

If you’ve enjoyed this post, or found it useful, please consider upgrading to a paid subscription for $4/month to make the newsletter sustainable, as it’s purely reader-supported. Or, check out my new book, FLUKE. I’ve been blown away by reviews so far, and I’ve immodestly included a smattering of them below from Amazon /Goodreads in case it might tantalize you a bit more. (Yes, I am a literary tease…but I’ve saved the best review for last).

Finally, there is also this excerpt from one particularly whimsical review. (If you hear that I’ve been captured in a net, now you’ll know what happened).

1. I think, anyway! I’ve not seen these terms used elsewhere, at least not in this way. But the phenomenon of multiple discovery means that someone else likely has before.

2. The United States has a very weird dynamic here, because having a news industry that’s fully representative of the population would also be a disaster, given that vast swaths of the public don’t think climate change/evolution is real, believe conspiracy theories, and/or hold racist/bigoted views. So, as with so many things in life, it’s a bit more complicated.