The Broken Window Fallacy, Broken Down
Why is the community worse off after the window is broken? What can the community do about it?
by Jonathan Newman
The Broken Window Fallacy
In a parable by Frédéric Bastiat and famously retold by Henry Hazlitt, a young hoodlum throws a rock through a baker’s window.
A crowd gathers to reflect on the incident and collectively concludes that although the window is broken, there is a bright side: a glazier will be hired to replace the window. The baker’s loss of a window is the glazier’s gain, and the glazier’s gain is the community’s gain because he can now spend his new income. This spending is somebody else’s income, and so on ad infinitum, so the broken window was the center of a rippling of benefit throughout the small economy.
It’s at this point Bastiat and Hazlitt say, “Wait a minute.”
The crowd forgot about someone: the tailor. You might glance back up to look for the tailor in my short summary.
He’s not there.
In fact, it is this aspect of the tailor (that he is unseen, hiding in the background) that makes this story not a story, but a parable with a lesson. The tailor comes into the light only when we consider the counterfactual, or what would have happened if the hoodlum had not thrown a rock through the baker’s window.
If the baker didn’t have to replace his window, he would have bought a new suit from the tailor with the money he handed to the glazier. So the community has not gained anything because of the hoodlum’s actions. It has actually lost something, namely, a new suit for the baker.
It’s at this point I say, “Did you catch that?”
Some may question the reasoning here. I’ll introduce some members of the crowd: Abigail, Bernard, Casey, and Daniel.
The Skeptics
Abigail pipes up first:
“Why has the town lost a new suit? Couldn’t the glazier have bought one with his new income? If not the glazier, what about the next people that receive the ripple of spending, or the next? Won’t the tailor sell a suit at some point?”
Bernard chimes in:
“Why don’t we let the government pay for the baker’s new window? It’s not the baker’s fault, so he shouldn’t have to pay — the government should, and that would cancel out the broken window.”
Casey gets a word in:
“How can we be sure that the baker wasn’t just going to hoard his money and not spend it at all? Isn’t the broken window a sure way to get it out of his pocket and into somebody else’s, stimulating spending?”
Daniel scratches his head:
“What if we just printed some money for the baker to use to pay for the new pane of glass? This way everybody benefits and nobody is worse off, including the baker! He can have his window and a suit and the glazier gets to start the same spending chain. It’s a win-win-win-win-win-win-win…”
Abigail interrupts: “We got it, Daniel. Now shut up.”
These are clever objections, but we’ll see how none of them disarm the Broken Window Fallacy, but commit some of their own errors instead.
But first, let me offer a definition:
Economic growth, defined
Economic growth is when consumers have the ability to satisfy more and higher-ranked ends.
“Gain”, “benefit”, and “better off” are all used in a similar sense and refer to which wants or desires we can satisfy and how many of our wants or desires we can satisfy. We single out consumers (and not, say, producers) for a couple of reasons: (1) everybody is a consumer, even producers when they’re not producing, and (2) all production is aimed at consumption.
Suppose Abigail wants a slice of pie and a cookie. The baker sells both for $5, and Abigail’ preference ranking (Abigail’s ends/wants/desires ranked by their importance to her) looks like this:
Abigail walks into the bakery with $5 in hand, ready for a treat. According to her preferences, $5 is too expensive for her to buy a cookie — she ranks the $5 over the cookie. The pie, however, is priced just right — she prefers the pie to the $5 and so gives up what she ranks lower in exchange for something she ranks higher.
This transaction has made Abigail better off. Since the baker also voluntarily exchanged the pie for $5, the baker is better off, too. The economy has grown because everybody involved is better off.
The blue items are what is given up. The orange items are what is gained. Green arrows mean things are moving up, which is indicative of all voluntary transactions.
What happened when the hoodlum broke the baker’s window?
Back to windows and suits, how can we map the broken window story using preference rankings?
Easy as pie.
The hoodlum likes breaking stuff. The baker likes his window, unbroken, but would also like a suit. The glazier is happy to replace windows for $500 and the tailor is happy to provide new suits for $500 (these more realistic 2015 prices are 10 times Hazlitt’s 1946 prices of $50 — apparently the powers that be didn’t read his other great book, What You Should Know About Inflation).
Then, the hoodlum does his thing:
He breaks the baker’s window, satisfying one of his ends, but at the detriment to the baker. The baker loses his most highly ranked end and gets something he regards as less than worthless in return, a pile of broken glass on the ground.
The baker turns around and uses his $500 to purchase a new window.
The baker and the glazier voluntarily exchange the $500 and the new window. The transaction is voluntary, even if the baker is miffed about the whole ordeal.
Meanwhile, the tailor is just sitting on the sidelines twiddling his thumbs.
The tailor never gives the baker a new suit. He only enters the picture when we consider the counterfactual, i.e., the alternate sequence of events had the hoodlum not committed his crime.
The counterfactual
What if the window were never broken?
The baker and the tailor trade one suit for $500. And everybody lives happily ever after. How happy, you ask? Well, compared to the scenario with the broken window, much better, which we’ll see in just a moment.
The two scenarios compared
A stocktaking of the two scenarios reveals that in the broken window scenario, the community has, in fact, lost a window. We’ll call the scenario that starts with the hoodlum breaking the window Scenario 1. The scenario without any vandalism is Scenario 2.
In Scenario 1, the baker loses his window at first and no amount of voluntary transactions after that can recover that loss.
We see that in Scenario 2, the suit enters the scene, and there’s nothing weighing down the “What is lost” side:
This type of comparison highlights the loss of the window in Scenario 1, but Bastiat and Hazlitt stressed the importance of the suit. When you just look at the “before” and “after” snapshots of Scenario 1, you see the town has lost a window. When you compare Scenario 1 to its counterfactual, Scenario 2, you see the that the town has also lost a suit.
Let’s compare what the baker walks away with after each scenario runs its course:
If the hoodlum breaks his window, the baker can’t get the suit. If the hoodlum doesn’t break his window, the baker can have both an unbroken window in his storefront and a new suit. He spends the $500 either way. The only difference is the suit, which only appears in Scenario 2.
Any chain of spending spurred by the glazier in the broken window scenario could also be spurred by the tailor in the scenario without a broken window. Therefore, anything beyond the baker must be considered the same for both scenarios.
Bastiat and Hazlitt conclude that the community is clearly better off in Scenario 2. Both represent a chain of spending, but one starts off with an “uneven” transaction, vandalism, while the other starts off with a mutually beneficial transaction.
Abigail’s objection
We are finally ready to take on the Skeptics’ objections.
Abigail spoke up first.
“Why has the town lost a new suit? Couldn’t the glazier have bought one with his new income? If not the glazier, what about the next people that receive the ripple of spending, or the next? Won’t the tailor sell a suit at some point?”
Abigail said that we can’t say for sure the town has lost a new suit because the glazier or somebody else could still buy the suit in the chain of transaction enabled by the replacement of the broken window. The baker may not get the suit, but somebody else could get it.
Let’s see what this looks like:
Compare this to:
So, Abigail is right in one sense. The choice of a suit as the opportunity cost of the broken window was arbitrary. Bastiat or Hazlitt could have chosen an encyclopedia set, a super large coffee drink, a trendy haircut, or any number of things the baker could have bought if he didn’t have to replace his window.
The point is, though, that the baker’s real wealth decreases as a result of the broken window. The wants he is able to satisfy are permanently fewer and lesser (lower ranked) than what he could have been able to satisfy without the vandalism.
No amount of spending, even in after an infinite number of transactions can make up for the initial loss the baker incurs. After an infinite number of transactions for both scenarios, Scenario 2 will still have the upper hand in terms of total benefit to everybody involved.
Bernard’s objection
Bernard felt bad for the baker having to pay for the vandal’s crime.
“Why don’t we let the government pay for the baker’s new window? It’s not the baker’s fault, so he shouldn’t have to pay — the government should, and that would cancel out the broken window.”
This objection is easier to dismiss. Taxing others to pay for the new window just shifts the harm to somebody else. If it’s unfair for the baker to pay for the new window, it’s also unfair for any other innocent person. The only person we can fairly place the burden on paying up is the vandal.
Bernard could respond:
“But we don’t like having people incur all the losses in a scenario like this. If we spread out the loss to a bunch of people, then situations like this don’t hurt as much.”
Bernard has stumbled onto something brilliant. We like knowing that if a tragedy like this were to happen to us, we wouldn’t have to suffer too much. The technical term for this phenomenon is loss aversion, and there are a few ways to deal with it:
- We could tax people and put the money in a fund for “broken window” situations. The tax, though, is applied to people who don’t agree with this program. Some people don’t want to be a part of this scheme, but the government takes their money anyways.
- We could have a market for insurance, where people voluntarily pay for coverage when “broken window” situations arise. Depending on how much coverage they want, they can pay more or less. The difference between this option and the previous one is that this one is completely voluntary. If somebody doesn’t want broken window insurance, they don’t have to pay in.
- We could ask for voluntary, charitable donations whenever these sorts of situations happen. The town bookseller, for instance, could ask people to donate to help pay for the baker’s new window.
The point is that we don’t need to apply more aggression by moving the loss onto people who aren’t involved and don’t want to pay in. Forcing people to buy into an insurance plan is just as unjustified as the vandal breaking the baker’s window because it involved taking something from somebody without their approval.
Casey’s objection
Casey was up next after Bernard.
“How can we be sure that the baker wasn’t just going to hoard his money and not spend it at all? Isn’t the broken window a sure way to get it out of his pocket and into somebody else’s, stimulating spending?”
This is another clever objection, but falls short when we consider the purpose of holding money. People save for future consumption.
Casey is partially right. We can’t be sure that the baker would have spent his $500 that day, or the next for that matter. He could wait years before spending it, in fact.
But a time delay in consumption does not eliminate the opportunity cost of the broken window. The baker’s preference ranking, according to Casey’s objection, could look like this:
With this arrangement, the baker would not spend his $500 today, but next year on a vacation. So the chain of spending would begin then, instead of today.
A market for loans in this economy would expand this line of reasoning even more. If there is a credit market in this small economy, then he could loan his $500 out to somebody else, maybe the barista so that she can buy a new espresso machine.
The espresso machine will allow the barista to sell more coffee drinks — if she made a good entrepreneurial judgment — and so she can pay the baker back a little more than the amount she borrowed. The baker’s savings, then, wouldn’t be a drain on the economy, but would help it grow by financing investment and increasing the amount this economy can produce.
Casey has another idea: “What if he never spends it? What if he goes mad and throws the $500 into the furnace that heats his oven?”
Even so, the baker would just be relinquishing his claim on $500 worth of goods, and would therefore increase the purchasing power of everybody else’s money. The baker would make everybody else better off at his own expense by burning the money.
Daniel’s objection
This leads us to Daniel’s objection about increasing the money supply.
“What if we just printed some money for the baker to use to pay for the new pane of glass? This way everybody benefits and nobody is worse off, including the baker! He can have his window and a suit and the glazier gets to start the same spending chain. It’s a win-win-win-win-win-win-win…”
It is true that in the short-run, things would look great. If the money printing authority in this small economy handed the baker some fresh, crisp bills — hot off the printing press — for him to pay for his new window, the baker could then have his window and a new suit.
But, just as a decrease in the small economy’s money supply made everybody else better off (like we saw in going over Casey’s objection), an increase has the opposite effect. The new $500 has introduced new claims on goods without actually increasing the amount of real goods in the economy. The money to goods ratio has increased and so prices will increase, too.
The following diagram is more complicated than the others, but the idea is simple. Printing new money and giving it to the baker allows for two potential chains of spending: one beginning with the transaction with the glazier and the other with the tailor.
Notice that since the goods are moving up in relation to the various dollar amounts, higher prices are charged for each good. This has two consequences: (1) people are paying more for the goods, and so have less to spend on other goods, and (2) it makes some transactions not happen at all because the price is now too high for one potential party to the transaction.
By the time the tailor gets the new money, prices have risen enough for him to value the encyclopedia set the same way the bookseller does: more than $510, but less than $520, so there is no opportunity for them to trade.
Even though I’ve “sped up” the effects of inflation so that we can see them in this small diagram, the mechanics are the same whenever there is an artificial increase in the money supply of any economy. The first people to spend the new money benefit at the expense of the last people to spend the new money.
Conclusion
Economic growth can only happen when people are allowed to make voluntary exchanges with each other.
In the Parable of the Broken Window as told by Bastiat and Hazlitt, the townsfolk were eager to find something good in a bad situation, but were looking in the wrong places. Involuntary destruction is always bad for an economy, and further distortions like taxes and monetary inflation exacerbate the problem. The best response to broken windows is to just let people work things out voluntarily.
If any more force should be applied after a vandal breaks a window, it should be used to make the vandal pay for fixing the destruction. Taxing others to pay for repairs makes innocent people pay for somebody else’s crime. Similarly, inflating the money supply forces innocent people to “pay” for the destruction in the form of higher prices or forgone benefit.
Broken windows are all around us and happen every day. Accidents happen, bad weather comes, and bad people steal and destroy our stuff. The counterfactual to broken windows is always preferable to broken windows, as Bastiat and Hazlitt explain, and there is no way to turn time back and stop broken windows before they happen. There is no way to recover the opportunity cost of broken windows.
Going forward, then, we should allow peaceful, voluntary exchange to persist and do what it does best: make people better off.
