The United States often touts itself as a capitalist economy, wherein each person operates according to their best interest. In theory, we are all mutually dependent, and people make money when they provide goods or services that are valued by other people – and for which they are willing to pay.

Adam Smith, writing generations ago in Wealth of Nations, coined the term “Invisible Hand,” which has haunted economic academia since he first put pen to paper. The Invisible Hand refers to the principle that markets are moved by the aggregate actions of individuals behaving in their own self-interest. The Invisible Hand would create markets for goods and services that are in demand; if a business would not meet the demand, it would close and another business would be created to meet that demand.

And that incentive model works well for certain things. If I invent a new kind of shoe that is more comfortable, or helps people dance better, or capitalizes on a new fashion trend, then I can make some and sell them, and make more money, and use that money to hire workers and build a shoe factory and make even more money and get rich. On the other hand, if the shoe that I make is not good, very few people will buy them, I will not make a lot of money, I will not build a factory and someone else will swoop in to fill the market’s appetite for shoes.

But not all markets work the same, and incentives are not always as straightforward as those that drive shoe sales. There are a few people alive today who can remember the passage of the Food and Drug Act. Congress felt that allowing the prescription drug market to operate on a pure incentive model – letting inventors sell drugs, and people deciding to buy more of the drugs that are “better” – was not acceptable. Congress believed that consumers needed more protection than a market could provide. Now, drug manufacturers are required to present evidence that a drug is “safe and effective” before selling it. It would be fair to argue that if a new drug resulted in people dying, consumers would stop buying that drug, but as a society, we are not willing to wait for people to die. So, we create regulatory bodies to protect consumers and oversee the drug market.

Another example is banks and insurance companies. Prior to the regulations of the 1930s, anyone putting money into a bank simply had to hope the bank was being a responsible steward of their deposit. Naturally, that system was unstable, and many people lost all of their savings when their banks failed. The problem is that bank managers have a huge conflict of interest. Say a bank owner decides that instead of using the money to make mortgage loans, they would gamble it all on one spin of the wheel at a casino. If they win, the bank has a huge profit and there are no consequences for their behavior. If they lose, the only money truly lost was the depositors’. The bank owner gets all of the upside if things go well, but only a fraction of the downside if their gamble goes poorly.

After the banking failures of the 1930s, it became clear that consumers needed protection, and indeed, that lasting damage could be done to the economic well-being of huge swaths of people if certain safety rails were not in place. Thus, we now have the Federal Deposit Insurance Corporation to insure consumers’ life savings will not be lost if a bank fails. And it imposes regulations to ensure bank executives cannot take enormous risks that threaten consumers’ well-being or the economy as a whole.

Another excellent example is airplanes. Airplanes are different from most other products, in that we have very high expectations for reliability. If 99.9999% of commercial airplane flights landed without crashing, that could still mean hundreds of people in the United States being killed in plane crashes every month. Obviously, that would be unacceptable. So allowing companies to sell airplanes that are reliable 99.999999% of the time while also allowing companies to sell planes that are reliable 99.9999% of the time is actually a profoundly dangerous mistake in this context. A market cannot favor one airplane over the other, because it would take years to distinguish those cases. Instead, we have robust and redundant safety systems that review the design of an airplane in excruciating detail before it is built and oversee its production.

But those systems depend on prioritizing safety. If the airplane is being made by people who are only in it for the money, planes will begin to become more unreliable. It turns out, safety is expensive.

And, in recent years, some decisions have been made that put cost-cutting ahead of safety.

One such safety measure is training by pilots on simulators for each type of airplane they fly before they actually fly passengers on that type of airplane. Low-cost airlines typically use a single type of aircraft so that they can switch pilots and airplanes whenever it is convenient. Most airplane disasters tend to be the product of overlapping and compounding failures. But one issue with the Boeing 737 MAX was that Boeing made major changes to other Boeing 737s while claiming it was an identical aircraft so as to avoid the requirement of additional pilot training. It’s impossible to know how things would have turned out differently, but it’s not hard to see how being upfront about the differences might have saved lives.

Corporations, just like communities or neighborhoods, have their own culture. Boeing was no different. The company originally manufactured airplanes and provided air transportation services. Most of their revenue came from the government, either providing airplanes for the military or, after World War I, providing air-mail services. After some scandals in both the Hoover and Roosevelt administrations in the late 1920s and early 1930s, the Airmail Act of 1934 required breaking up the company, so it became Boeing, the airplane manufacturer, United Airlines the airline, and United Technologies.

Boeing built a reputation for manufacturing safe airplanes, investigating every disaster to find the root cause and preventing it from happening again.

But in the 1990s, after Boeing merged with McDonnell Douglas, the company changed from being led by engineers to being led by their finance department. Over time, they switched from building the best airplanes that they could afford to build, to building the most profitable airplanes that would meet the letter of the law on safety and reliability.

The invisible hand cannot deal with this issue because:

— There are only a few companies in the world that make airplanes. Two companies (Boeing and Airbus) make almost every large commercial airplane used by airlines.

— Designing a large airplane is such a huge project that it would take years for a new company to start delivering airplanes.

In fact, there are a lot of things for which the invisible hand does not work in an acceptable way.

— We have decided that restaurants must have certain governmentally enforced safety standards, because we don’t want to wait for people to get sick to find out that a restaurant is not operating safely.

— We have decided that drugs must be safe and effective.

— We have decided that automobiles must meet many safety standards.

— We have decided that insurance companies must follow many rules and regulations.