Bastiat's Broken Window Is the Best Economics Lesson in Five Minutes

A boy throws a rock through a baker's front window. Glass everywhere. The baker is furious, the crowd gathers, and the boy is marched home. Then somebody in the crowd, a person with the air of a man who has taken one economics course, says the thing. "Well, look at it this way. The glazier will get work. He'll buy bread, the baker will have more money, and it all circulates. Maybe the boy did the town a favor."

Everyone nods slowly. And everyone is wrong. The mistake in that sentence is the most expensive mistake in public life, and a French writer spotted it in 1850. I want to give it to you in five minutes, because once you see it you will spot it in every press conference for the rest of your life.

The writer was Frederic Bastiat, a Frenchman who served in his country's legislature, wrote with a wit that economists rarely manage, and died in Rome in December 1850 at the age of forty-nine. A few months earlier he had published a short essay with a title that is the entire lesson: That Which Is Seen and That Which Is Not Seen. It is still the best thing ever written on how to think about money in public life.

The parable, step by step

Bastiat's version has a shopkeeper whose careless son breaks a pane of glass. Onlookers console the father by saying that accidents are good for trade: the glazier gets paid, and the glazier will spend that money, and so on. Bastiat's response is to follow the money all the way to the end. I will use dollars instead of francs.

Step one: the seen. The window is broken. A glazier charges $600 to fix it. The shopkeeper pays. The glazier now has $600 he did not have, and we can see him walking away with it. This is the part the crowd noticed.

Step two: the unseen. The shopkeeper had $600 in his pocket before the rock flew. He had a plan for it. Suppose he meant to buy a new suit. Now he cannot, because the money went to the glazier. The tailor does not get the $600. Nobody sees the suit that was never made, never sold and never worn, because it does not exist.

Step three: do the books. Before the accident, the town had a good window and a shopkeeper with $600 to spend. After the accident, the town has the same good window, and the shopkeeper has no $600. The glazier is $600 richer, and the tailor is $600 poorer. Net: the town has the same window it started with, and it is short one suit.

Step four: state the lesson. The broken window did not create $600 of economic activity. It moved $600 of activity from the tailor to the glazier and left the town one suit poorer than it would have been. The glazier is the seen. The tailor is the unseen. Bastiat's rule is that a good economist counts both and a bad economist counts only the one he can see. The wealth of a society is not the amount of activity in it. It is what it has to show for the activity.

Henry Hazlitt, who made the same point famous in America, said the whole of economics can be reduced to one lesson: look not only at the immediate effects of a policy but at the longer effects, and at the effects on all groups, not just the favored one. He put it in a book, Economics in One Lesson, published in 1946, and he began it with a chapter titled "The Broken Window."

Apply it to taxes

Now take the same method into the real world. Suppose, as a hypothetical, a town of 1,000 families levies a new tax of $200 on each family to fund a program.

The seen: $200 x 1,000 = $200,000. The program uses it to hire four workers at $50,000 each (4 x $50,000 = $200,000). Four people have jobs, there is a ribbon-cutting, and the local paper runs a photograph.

The unseen: Each of the 1,000 families has $200 less. That is $200,000 of spending, in the aggregate, that does not happen: the restaurant meals, the kids' shoes, the car repair, the savings that would have become a loan to a local business. Those purchases would have supported roughly the same number of jobs, spread thinly among the hardware store, the diner and the mechanic, so thinly that none of them can point to the one job that was lost.

So the program did not create four jobs. It moved four jobs from where families would have chosen to put them to where the government chose to put them. The question is whether the four new jobs produce something worth more to the town than what the families would have bought. Sometimes they do. A police officer, a road crew, a teacher may well be worth more than a restaurant meal. But the test has to be made honestly, with both columns on the page, and not by pointing at the jobs we can photograph.

I want you to notice that this is not an argument that all public spending is bad. It is an argument for counting. A program that passes the test has earned its tax. One that cannot be defended except by "look how many jobs it created" has failed it. The checklist in Ten Questions to Ask Before You Support Any New Government Program puts the same test into ten questions.

Apply it to subsidies

A subsidy is the seen and the unseen in a pure form. When the government pays a company to build a plant, we see the plant, the workers and the ribbon. What we do not see is what the same capital would have done if it had stayed with the taxpayers who supplied it, or with the other business owners who did not get a subsidy and had to compete with one that did.

Take a hypothetical state that offers $50 million to land a factory with 500 jobs. The seen is 500 jobs. The arithmetic of the unseen starts with $50,000,000 divided by 500, which is $100,000 of public money for every job. If the plant would have come anyway, the subsidy bought nothing. If it would not have come, the cost is still real, and it was paid by everyone who was taxed and by the competing businesses that got no help. Ask what the $50 million would have done if left in private hands. That is the unseen column.

Our own side deserves this lesson as much as anyone. Farm subsidies, sugar quotas, ethanol mandates and tax breaks for favored industries have been defended by politicians of both parties, because the beneficiaries are visible and organized and the costs are scattered. See zoning may be the biggest price control you never got to vote on for one of the hidden ones, and Seven Myths About the Free Market That Deserve Retirement for a broader tour.

Apply it to make-work

Bastiat had great fun with this one. In his 1845 satire, the Candlemakers' Petition, a fictional group of candlemakers asks the government to pass a law closing every window, shutter and curtain, so that the sun would stop competing with their product. Think of the jobs, they say. Think of the candles, the wax, the wicks, the whale oil.

The joke is that the argument is identical to the one made, in earnest, for every protective tariff and every rule that forces a business to use more labor than it needs. If jobs were the goal, we could employ everyone by banning the tractor and handing out spoons. Total output would collapse. The point of an economy is not to maximize effort. It is to produce goods and services that people want, with as little effort as possible, so that the effort saved can make something else.

That is why economists ask not "how many jobs did it create?" but "what did it cost per job, and what did we get?" And it is why the argument for who really pays a tariff, following the dollar from dock to shelf is often an argument about the unseen. The visible job is in the protected factory. The invisible cost is spread among the shoppers.

Two more of Bastiat's own examples

The window is the famous one, but the same essay works through several others, and two of them are useful at a town meeting.

Public works. Bastiat took on the argument that a government project, such as a canal or a road, puts people to work. The seen is the canal and the crew digging it. The unseen is the work those same tax dollars would have paid for if left with the people who earned them: a family's new roof, a farmer's new plow, a shop's new sign. The test is whether the canal is worth more than the sum of those private uses. Some canals are. The point is that "it employs people" is never the answer, because anything employs people. Lighting a pile of money on fire would employ a fire marshal.

Machinery. In the same spirit he took up the fear that new machines destroy jobs. The seen is the weaver put out of work by the loom. The unseen is that cloth now costs less, so families have money left over to spend on something else, and that new spending calls for new workers somewhere. The transition is painful for the weaver, and Bastiat did not pretend otherwise, but the nation that bans the loom is a nation of people who pay more for cloth and have less left for everything else. Americans have run this experiment many times, from the plow to the assembly line to the computer. The towns that tried to freeze the old way of working rarely held on to the jobs, and they lost the lower prices as well.

Both examples teach the same discipline. Train yourself to ask "and where did that come from?" every time somebody describes a benefit. It is the question that separates an economist from a salesman.

The limits of the parable

An honest teacher has to say where the lesson stops. Here are the best objections.

One: idle resources. The strongest objection comes from the tradition of John Maynard Keynes. In a deep downturn, the argument goes, there are idle workers and idle factories, and money sitting in bank accounts that nobody will spend. If the shopkeeper would not have bought the suit anyway, because he was afraid and saving every dollar, then the tailor was never going to get the money, and the glazier's work really is a net addition. In that world the unseen is not a suit. It is nothing. Economists still argue over how often the world looks like that, and how fast the benefits of public spending arrive compared with its costs. A conservative should grant that the objection is serious and then point out that Bastiat's lesson applies with full force in the normal case, and that the case for stimulus has to be made on its merits, not by saying that destruction is productive.

Two: the unseen can be a benefit too. The method works in both directions. A bridge built with taxes has an unseen side: the trade that crosses it, the time saved, the lives that are not lost in a ferry accident. Counting the unseen means counting benefits that nobody photographed. A road, a court system and an army pass because their unseen benefits are very large. The parable does not say "never spend." It says "count everything."

Three: it does not tell you what people should want. Bastiat's test is about cost and benefit as people value them. It cannot settle whether a society should spend on a museum or on a stadium. That is a question of values, to be settled in the open by voters.

Four: disasters really do bring rebuilding. After a hurricane, rebuilding creates work and shows up in the numbers. The parable's answer is that the numbers measure activity, not wealth. The town was richer before the storm. Rebuilding is better than doing nothing, but it is not better than not having the storm. A person who calls the storm a boost to the economy has made Bastiat's mistake.

The follow-up: Hazlitt

If you read one book after this post, read Henry Hazlitt's Economics in One Lesson. Published in 1946, it takes Bastiat's method and runs it through taxes, public works, tariffs, minimum wages, price controls and credit. It is short and clear, and it has the right enemy: the habit of looking only at what is in front of you. Hazlitt was a journalist, not a professor, and wrote for people who pay bills. He also gives you a good habit for your own use: whenever someone proposes a policy, ask who benefits first, then ask who pays, then ask what the money would otherwise have done.

If you want the economics behind prices, pair it with Hayek Explained Why the Smartest Planner in the Room Always Loses, and for the principle that markets coordinate millions of decisions without a coordinator, I, Pencil and the price system, how nobody plans your breakfast. Both reach the same destination by different roads: nobody can see everything, so be humble about the part that you can't.

The bottom line

• Every cost has two sides: what you see (the program, the plant, the glazier) and what you do not (what the money would have done, the tailor). Count both columns before you cheer.
• A tax of $200 on each of 1,000 families takes $200,000 out of the town's private spending. If a politician says it "creates four jobs," ask which jobs it destroyed.
• Be as hard on your own side's favorites as on the other side's. Tariffs, quotas and tax breaks for favored industries have the same structure as the broken window.
• Grant the limits. In a deep slump with idle workers and idle factories, the unseen may be small, and public spending has a better case. Make that argument openly instead of hiding behind "jobs."
• Read Bastiat's essay, all of it, and then Hazlitt. Five minutes with the first and an evening with the second will make you the hardest person in your town to fool.