One day a boy was playing football when he accidently broke a window. Rather than get mad, the people shrugged their shoulders and said breaking windows is good for the economy. After all, if no windows were broken, then all the glaziers would be out of a job. By breaking the window, the boy ensured money would be spent on repairs, thereby ensuring someone kept their job and giving the window making business a boost. However, at this point Bastist in his seminal essay “That Which Is Seen And That Which Is Not Seen” jumps in to point out why this is a fallacy. While we see the money spent on repairing the window, we don’t see what would have happened had the window not been broken. Instead of repairing the window, the money could have been spent on a new pair of shoes. So while the glazier is better off, we don’t see the people who are worse off as a result.
The first time I heard this, I thought it was a nice little parable from long ago and not much more. It is obvious that destructive activity is not productive and this was a nice way of saying it that didn’t quite re-invent the wheel. After all broken windows and natural disasters are rare even in America (and almost non-existent here in Europe) and have very little economic effect. However, to my surprise, some people thought this was a hugely important lesson. Libertarian blogs such as Mises and Cafe Hayek extolled the virtues of the parable as of huge importance for understanding economic policy making today. Probably the most famous libertarian economic book called “Economics In One Lesson” is dedicated to explaining this very fallacy. In fact it is widely seen as a silver bullet that completely defeats Keynesian economics altogether. How?
It is argued that when we see the government spending money on a public works scheme and think we see jobs being created, we are just like the people who think that broken windows boost economic activity. We are ignoring the unseen effect. While we see the new road and the workers employed on it, we do not see what would have happened had the road not been built. It is argued that the government no more creates jobs than breaking windows creates jobs. In order to get the money to pay for the road, the government must impose a tax which takes money that consumers would have spent elsewhere. It also takes employees who would have worked elsewhere and materials that could have been used on other jobs. In other words any increase in government activity is offset by a decrease in private sector activity.
This is an interesting argument as it assumes that 1) Say’s Law holds and 2) the economy is at full capacity, which means that there is 100% crowding out. What is interesting about these theories is that they were dominant in the 19th century (when this essay was written) but were debunked during the Great Depression. Not even mainstream neo-classical economics textbooks teach Say’s Law anymore and it is rarely heard of. So in a sense the Broken Window Fallacy is like rereading arguments over Irish independence or speeches from the Civil War or the battle lines of World War One. It is offering an insight in a historical period long past and a glimpse at how people used to think. Unfortunately some people go a bit too far in their historical re-enactment and think the Broken Window Fallacy has some relevance for today.
The strange thing about assuming full capacity as the Broken Window Fallacy is that you end up in a strangely paralytic nihilistic world. If any action by the government is automatically offset, then surely the same can be said for the private sector? If I set up a restaurant, some might praise me for being an innovative entrepreneur, but am I not depriving the rest of the private sector my rental space? Is there not one less chef for other restaurants to hire? If I am praised solely on the money I spend, are we not forgetting about what is unseen? After all, my money didn’t come from nowhere, my business may be richer, but are my customers not equally poorer? For every customer I gain, does that not mean all other restaurants have one less customer? Is all investment, public or private, not futile?
The perverse conclusions from 100% crowding out do not end there. If there is full capacity then government intervention can be justified. A friend once told me that the economy needs people to smoke due to the boost that cigarettes give in terms of employment and taxes. But if there is full capacity then even if cigarettes were banned tomorrow people would have extra money in their pocket. Other businesses would get a boost equal to the decrease of the tobacco industry and the unemployed workers could be rehired. Thus if the government strongly discourages one industry this does no overall damage to the economy but merely shifts resources from one area to another. So if economic output is the same regardless, then the government can easily intervene to pick the right industry and discourage others like alcohol and tobacco.
The problem of course is that we are in a time of recession when we are far below full capacity and there are idle resources. So if someone spends money now, there is no guarantee that they would have spent it anyway. When a person is put to work this does not mean they would worked somewhere else otherwise, there is a good chance they would be have been unemployed. So the choice is not always between government spending or private spending, but government spending or no spending. But some would argue, if people don’t spend, they save which means the money will eventually get spent in the long run. But that is little good to us now (this was the key point of Keynes most famous and most misunderstood quote). When people are trapped in unemployment, it is not much good to tell them that in three years time there will probably be jobs. We are in a crisis now and we need a solution now before the recession does long term damage to potential output. The point of a stimulus is to speed up the process and to borrow from the sunny future to pay for a rainy today.
Seen and unseen is an interesting point and one that can be applied to libertarians too. When a libertarian decries the minimum wage, they are only focusing on the costs the employers’ face, that is what is seen. However, they fail to notice the unseen, namely that as employees wages are higher, they spend more, hence sales and revenues of a business rises. This is why “minimum wage= unemployment” is a statement asserted rather than a fact proven.
The fallacy assumes that the displaced private sector activity was of equal or greater value to the new public investment. But what if the activity had a negative effect on society? For example what if the government taxed pollution or alcohol to pay for the public works? In this case a reduction in private sector activity caused by crowding out is actually a good thing. A Pigouvian tax on tobacco, alcohol, carbon etc would still be beneficial to the economy even if the Broken Window Fallacy holds. It is also incorrect to assume that private and public sector spending are the same. So even if a new public hospital built by a tax on luxury cars does not increase GDP, it can still be argued that society is better off. Wealth can still be redistributed even if output is not increased.
Keynesian do not claim that breaking windows creates wealth, rather they show that even something as destructive as breaking windows can increase employment and boost economic growth. Critics go wrong by taking Keynesians too literally. The point of broken windows is not that this is a great idea we should all implement, but that if even destructive activities can boost the economy, imagine how could can come from investments that are desirable in their own right such as schools and hospitals. If breaking windows can boost the economy and building wind turbines can prevent climate change imagine what happens when you combine the two? Breaking windows isn’t the core teaching of Keynesians, but rather an example of even in the worst case scenario, the economy gets a boost, so imagine what would happen in the best case.