The “Expert Fallacy”
When you’ve got a problem you’ve never faced before, you find an expert.
If you have a software problem, you find a developer. If you’re a political leader with a potential pandemic, you find an epidemiologist. If you have a strategic analysis problem, you find someone like me.
You find an expert.
But not every expert is a good expert.
Because not every expert is willing to recognize the limit of their expertise.
In the spy world, you see an expert in tribal conflicts try to make decisions about economic issues. You see a manager tell an officer how to deal with a source the manager has never met (In A Spy’s Guide To Strategy is a story about that). You see experts overstep their bounds. You see them make assertions and talk confidently about things they don’t know anything about.
And the amazing thing is: People listen.
People listen to experts, even when the expert is talking about things outside their expertise.
So many people do it, psychologists have given it a name: The Expert Fallacy.
The Expert Fallacy is when we trust an expert when the expert is talking about things outside their expertise.
It looks like this:
Some try to get around that problem by trusting lots of experts. Find an expert on every possible topic. Every possible problem. Every possible thing that could happen.Some try to get around that problem by trusting lots of experts. Find an expert on every possible topic. Every possible problem. Every possible thing that could happen.
They try to make the world look like this:
Instead of one expert telling us what to do, we trust a group of experts. Because the group, together, can answer all the questions we have, from epidemiology to economics to environmental issues.
But even with a group of experts, there’s still the chance of the Expert Fallacy.
Because there’s still the chance that an individual expert will talk about things outside their expertise.
Sometimes, it happens innocently. A journalist asks the expert a question they don’t know the answer to. To move the conversation along and please the journalist, the expert guesses at the answer. Their guess is just as good as anyone else’s. But because they’re an expert, it’s given more weight. The journalist repeats the expert’s guess, and lots of people believe it’s true. We end up back at the Expert Fallacy.
Or, the expert purposefully expounds on things outside their expertise. Because it’s rewarded by people who like their answers.
Or, the expert gets frustrated with all the bad decisions being made based on the expert’s data collection and analysis. The expert thinks they could do a better job making decisions than the decision-makers, so they move out of their area of expertise. They want to go to the right in the D-A-D-A sequence (graphic from A Spy’s Guide To Thinking):
Sometimes, we want that. Sometimes, we want experts to be the decision-makers. But history has shown us that letting experts make big decisions can be dangerous.
We could talk about the many times experts have been wrong on big questions (the fall of the Soviet Union, smoking’s connection to lung cancer, almost every US political question of the last four years), but better is to point to an interesting societal evolutionary result.
We don’t let experts decide what is arguably the most important question in society: Whether someone is guilty or innocent of a crime. Instead of experts, we let 12 random people on a jury decide.
The 12 random people on a jury listen to lots of experts. Experts in what happened (witnesses). Experts in evidence (crime scene investigators). Even experts in accounting, sometimes.
But those experts aren’t the decision-makers. They don’t decide whether someone is guilty. That’s up to the jury.
Why is that? Why do we trust 12 random people over experts?
It’s because lots of alternatives have been tried. Way back when, we tried letting chance decide (drawing straws). Then we tried letting those with political power decide (the king is the law). Then, we tried letting experts decide (for example, the Spanish Inquisition).
As a civilization, we didn’t like any of those ways of determining guilt more than trial by jury.
Trial by jury isn’t perfect, but it’s better than the alternatives (as the victims of the Spanish Inquisition would tell you).
You don’t want experts deciding whether you’re guilty or innocent. And we don’t want experts deciding other big questions.
But we still need experts for lots of things. We need experts to collect data, analyze it and provide answers to the decision-makers.
We need experts to help us make decisions.
How do we make sure experts aren’t outside of their area of expertise?
How do we avoid the Expert Fallacy today?
How to judge experts
In the Spy’s Guide series are several mental models to help you work through important questions on the fly.
The most general of those models is the D-A-D-A model in A Spy’s Guide To Thinking. It’s a simple model from the world of intelligence to help you focus on the right things to get to answers quickly. So you can decide and act.
But it can’t get you to the right strategic questions to ask. That’s why there’s A Spy’s Guide To Strategy, which has a simple model to help you figure out the strategic questions to ask. And also A Spy’s Guide To Taking Risks, which has a simple model for managing risks.
And these mental models also work for judging experts.
If you go back and look at these books again, you’ll find:
The D-A-D-A model helps you zero in on the data that underpins the expert’s analysis. It helps you ask the right questions about the expert’s analysis. The game theoretic model in A Spy’s Guide To Strategy helps you figure out whether the expert is playing a positive-sum or zero-sum game with you. And the risk model in A Spy’s Guide To Taking Risks helps you judge how likely success is from following the expert’s advice.
In this world, we’re all decision-makers.
And one of the most important decisions is always:
Which experts should we trust?
— — — — —
My thanks to all who continue to recommend the Spy’s Guide series to friends. Five years after the first book was published, it’s still on the Kindle Nonfiction Single bestseller list. Many thanks!
