He had help with this too.

Photographer: Yuriko Nakao

# Did Elon Musk Build That?

Noah Smith is a Bloomberg View columnist. He was an assistant professor of finance at Stony Brook University, and he blogs at Noahpinion.
a | A

Do guns kill people, or do people kill people? This question points to a cognitive error that humans make all too often. The error is in the way we think about the causes of things, likening causes to weights on a scale.

When you have weights on a scale -- or on a barbell at the gym -- you can find the total amount of weight just by adding them up. Ten pounds plus five pounds equals 15 pounds. People tend to think of causation the same way. We ask how much personal or professional success is due to nature and how much is due to nurture. We ask how much of global warming is caused by humans and how much is driven by natural factors. Basically, we are looking at a barbell, and asking how much of its weight comes from each metal disc.

For some applications, that's fine and good. For others, though, it's terrible. Take gun crime. How much is a result of the easy availability of guns, and how much owes to human behavior? To have a gun murder, you need both a gun and a person willing to pull the trigger. Take away either, and you no longer have a gun murder. So if we use the "weights on a scale" method, we reach a nonsensical conclusion -- we end up saying that both guns and people cause 100 percent of gun crime, which adds up to more than 100 percent.

In cases like this, it's better to think of causality as a chain, rather than weights on a scale. When a chain pulls, you need every link to hold, or else the chain snaps. Asking "Which link is doing more of the pulling?" doesn't make sense.

Actually, these two concepts of causality have a name in economics -- they're called "substitutability" and "complementarity." Substitutes are like pens and pencils -- if you have less of one, you can make up for it with more of the other. Complements are like left shoes and right shoes -- you need one of each to make a pair.

Humans often think of causes in terms of substitutes -- as weights on a scale -- but most processes involve some degree of complementarity. This cognitive error can poison our policy discussions.

For example, take the discussion over whether companies are built by their founders or by institutions. People on the left tend to downplay the contributions of founders -- for example, in the MIT Technology Review, Amanda Schaffer writes:

By warping the popular understanding of how technologies develop, great-man myths threaten to undermine the structure that is actually necessary for future innovations...

[Elon] Musk’s companies...rely on public-sector support and good timing...“SpaceX is surfing on years and years of government-funded technology and public-sector support,” as Mariana Mazzucato, an economist at the University of Sussex and author of The Entrepreneurial State, points out...

Likewise, Musk’s success at Tesla is undergirded by public-sector investment and political support for clean tech.

Schaffer implicitly frames the "great man" theory as saying that institutions don't matter. But the fact is, both matter 100 percent. Entrepreneurs and institutions are highly complementary. Without the government research that created space technology, Elon Musk's SpaceX wouldn't have been created. But without Elon Musk, it also wouldn't have been created. Musk and the government are two links in a chain, each of which is indispensable, not two weights on a scale.

This is why so many people misunderstood President Barack Obama's "you didn't build that" speech. Obama was trying to explain that the government was an essential link in a chain of causality -- that if you neglect good institutions, business founders will find it very hard to do what they do. But because people have trouble thinking about causal complementarity, they thought Obama was downplaying the contributions of entrepreneurs. A causal chain was too hard even for a great communicator like Obama to explain, because we're just so used to thinking we can add causes up like weights.

So how should we think about causal chains? If we want to encourage something -- such as business startups -- we should think about the weakest link in the chain, and how to make it stronger. If we find that, like Japan, we have lots of good technology and business resources, but few entrepreneurs, we should think about encouraging people to take the risk of starting their own companies. But if we find that, like most poor countries, we have a surplus of entrepreneurs but weak institutions to back them up, we should strengthen things like property law, contract enforcement or research funding.

So even though it’s difficult and can feel unnatural, we should try our best to understand causality as a chain. It would significantly improve a lot of our public discussions.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

To contact the author on this story:
Noah Smith at nsmith150@bloomberg.net

To contact the editor on this story:
James Greiff at jgreiff@bloomberg.net