GDP and Broken Windows

[Autolykos]

In light of recent events, I’d like to explain Bastiat’s parable of the broken window in a more, shall we say, econometric format.

Using the standard definition of GDP (“the total market value of all goods and services produced within an area in a given period”), we can measure the GDP of a fictional village in France called Parableville. In this village, James Goodfellow’s window has just been broken by his careless son. He will now pay the glazier six francs to get his window replaced. This expenditure is added to Parableville’s GDP. Assuming the GDP for the previous period was 100 francs, then the GDP for this period (ceteris paribus) will now be 106 francs. That’s an increase of 6% in GDP for Parableville. Clearly breaking windows can lead to increased GDP.

However, as Bastiat noted, there’s what is seen and there’s what is unseen. What’s unseen here? James Goodfellow’s loss, of course. Since he paid six francs to the glazier to make a new window for him, we can estimate the current market value of his old window as also being six francs. If we offset this loss against the new window, we find that (again ceteris paribus) the GDP for Parableville stays at 100 francs. In other words, when we take economic losses into account, breaking windows doesn’t lead to increased GDP.

There’s more to this story, though. Let’s now say that, had James Goodfellow’s window not been broken, he would have paid the cobbler for a new pair of shoes. Let’s further suppose that James would have paid the cobbler six francs for this. Doing so would also raise the GDP for Parableville to 106 francs (again ceteris paribus), but this time there’s no accompanying loss, so the GDP value stays there. Compared to the “seen” GDP, this “unseen” GDP is 6% higher. Inverting this, we can say that breaking the window resulted in a “seen” GDP figure that is 6% lower than the “unseen” figure.

Now let’s change the scenario to something a bit more modern. Let’s say that the government of Parableville takes six francs from James Goodfellow in taxes. It then spends that six francs on a new window from the glazier. Following the definition of GDP, this would raise Parableville’s GDP from 100 francs to 106 francs (ceteris parabus). Once again, however, this doesn’t take into account the loss against James Goodfellow, which is also six francs (the amount he was taxed). Offsetting this loss, and holding to ceteris paribus, brings the GDP back down to 100 francs.

Had the government of Parableville not taxed those six francs from James Goodfellow, he would have used them to pay for a new pair of shoes from the cobbler, as before. Once again, there would have been no offsetting loss here, so Parableville’s GDP would have genuinely grown by six francs, or 6% over the previous period’s GDP value of 100 francs (again ceteris paribus). The inverse of this is again that taxing six francs from James Goodfellow resulted in a GDP figure that’s 6% lower than what it otherwise would have been.

So we can see that taking into account offsetting losses would be a significant step in making GDP more accurate. Not only that, but doing so could also give us at least an idea of how much more the GDP would have grown had those offsetting losses not occurred. To sum up, the GDP measure as used by econometricians the world over in no way accounts for all of the effects of “broken windows”.

[Wheylous]

There are a couple flaws in your scenarios which I believe make them unworkable.

First of all, GDP is only produced goods. When something is destroyed, it does not count toward production. There might be some loss of total wealth, but there is no destruction of GDP.

Furthermore, the loss of the window is not necessarily a loss of 6 francs. If the window was old and slightly cracked and covered in little smudges of paint, it would not be worth 6 francs, especially given the improvements in technology since the creation of the window which will lead to better windows being made for the same price.

The point is, the window is not on the market, and so there is no “cost” of the window. The cost of something is what you give up to get the item. Yet since there was no market for the window, there was no cost for the window. Trying to add it in is making up fictitious value for this and all aforementioned factors.

Moving on,

Even assuming that GDP somehow had to account for losses, your example leads to some “interesting” conclusions. If breaking the window to replace it is a loss of 6 francs (ignoring the arguments above), then why can’t we say that the same happens when a company buys a house, demolishes it, and builds a factory on it. Surely, the house must be subtracted from you GDP-like measure.

Not only this, but things like weathering of buildings and roads will have to be taken into account.

At this point, your GDP becomes a measure of wealth, not of production per year.

Now let’s change the scenario to something a bit more modern. Let’s say that the government of Parableville takes six francs from James Goodfellow in taxes. It then spends that six francs on a new window from the glazier. Following the definition of GDP, this would raise Parableville’s GDP from 100 francs to 106 francs (ceteris parabus). Once again, however, this doesn’t take into account the loss against James Goodfellow, which is also six francs (the amount he was taxed). Offsetting this loss, and holding to ceteris paribus, brings the GDP back down to 100 francs.

Cash transfers do not affect GDP. If the money is taken from A and given to B, this doesn’t change GDP. Now, the government spends the money and this increases GDP, yes. But 1) We’re not guaranteed that James would have spent the money (perhaps he is a saver) and 2) This doesn’t “bring down” the GDP. We’re merely saying that the government spent the GDP rather than James.

I’m afraid your analysis is unworkable. Both because you’re twisting what GDP is trying to measure and because you’re somehow saying that when the government pays for production this isn’t really production.

[Wheylous]

To explain why your “fake government GDP” is wrong:

Pure market scenario:

Private consumption of town: 100 francs

Tax from James: 0 francs

James then buys shoes: 6 francs (adds 6 francs to GDP)

Total GDP: 106 francs

Government taxation and spending:

Private consumption of town: 100 francs

Tax from James: 6 francs (doesn’t affect GDP)

Government then buys additional window: 6 francs (adds 6 francs to GDP)

Total GDP: 106 francs

While you could try to argue that the 6 francs the government spent are inefficient allocation of resources, this doesn’t change the fact that the sum total of goods produced in the economy sold for 106 francs for the period in both cases.

[Neodoxy]

Autokylos I’m sorry but I really don’t know what you’re trying to say. Just as Wheylous states, the fact that the government will spends the tax money means that GDP is, at least, kept constant. Indeed the transfer is likely to increase GDP specifically because people tend to save much more than the government does, which is part of the problem.
The significance of Bastiat’s fallacy is twofold:

1. It shows that even if people aren’t forced to spend their money on one thing that they are likely to spend at least part of it, and therefore nothing would be waisted (I’d argue that part of Bastiat’s failure is that he didn’t properly take into account the consequences of savings).

2. That simply spending does not mean an increased in happiness. Even if GDP did decrease as a result of the window not being replaced then wouldn’t people still be better off if they only spent half the money on other things? Of course, a large part of this depends upon price flexibility and the multiplier.

[Autolykos]

@Wheylous

First of all, GDP is only produced goods. When something is destroyed, it does not count toward production. There might be some loss of total wealth, but there is no destruction of GDP.

I’m aware of that. I thought I made that clear in my post.

@Wheylous

Furthermore, the loss of the window is not necessarily a loss of 6 francs. If the window was old and slightly cracked and covered in little smudges of paint, it would not be worth 6 francs, especially given the improvements in technology since the creation of the window which will lead to better windows being made for the same price.

The point is, the window is not on the market, and so there is no “cost” of the window. The cost of something is what you give up to get the item. Yet since there was no market for the window, there was no cost for the window. Trying to add it in is making up fictitious value for this and all aforementioned factors.

Of course it’s not necessarily a loss of 6 francs. Nothing has inherent value. My point was not to try to rehabilitate GDP, it was just to illustrate how much GDP misses.

@Wheylous

Moving on,

Even assuming that GDP somehow had to account for losses, your example leads to some “interesting” conclusions. If breaking the window to replace it is a loss of 6 francs (ignoring the arguments above), then why can’t we say that the same happens when a company buys a house, demolishes it, and builds a factory on it. Surely, the house must be subtracted from you GDP-like measure.

Not only this, but things like weathering of buildings and roads will have to be taken into account.

At this point, your GDP becomes a measure of wealth, not of production per year.

Indeed, which highlights how inaccurate GDP is, if you ask me.

@Wheylous

Cash transfers do not affect GDP. If the money is taken from A and given to B, this doesn’t change GDP. Now, the government spends the money and this increases GDP, yes. But 1) We’re not guaranteed that James would have spent the money (perhaps he is a saver) and 2) This doesn’t “bring down” the GDP. We’re merely saying that the government spent the GDP rather than James.

I understand that cash transfers don’t affect GDP as it is. That was again part of my point. I’m sorry if it wasn’t clear enough - I certainly tried to make it as clear as possible.

@Wheylous

I’m afraid your analysis is unworkable. Both because you’re twisting what GDP is trying to measure and because you’re somehow saying that when the government pays for production this isn’t really production.

True, I am twisting what GDP is trying to measure, the point of which is to highlight how much of the picture I think GDP misses. When the government pays for production, it is production, just as it is when the hypothetical James Goodfellow pays the glazier for replacing his broken window. How is my analysis unworkable again?

[Autolykos]

@Wheylous

To explain why your “fake government GDP” is wrong:

Pure market scenario:

Private consumption of town: 100 francs

Tax from James: 0 francs

James then buys shoes: 6 francs (adds 6 francs to GDP)

Total GDP: 106 francs

Government taxation and spending:

Private consumption of town: 100 francs

Tax from James: 6 francs (doesn’t affect GDP)

Government then buys additional window: 6 francs (adds 6 francs to GDP)

Total GDP: 106 francs

While you could try to argue that the 6 francs the government spent are inefficient allocation of resources, this doesn’t change the fact that the sum total of goods produced in the economy sold for 106 francs for the period in both cases.

Again, I know that the standard GDP measure would say that. I don’t see how I denied that anywhere in the OP. My point was to show what’s missing from the GDP measurement. I don’t believe I’m mistaken about how GDP actually works.

[Autolykos]

@Neodoxy

Autokylos I’m sorry but I really don’t know what you’re trying to say. Just as Wheylous states, the fact that the government will spends the tax money means that GDP is, at least, kept constant. Indeed the transfer is likely to increase GDP specifically because people tend to save much more than the government does, which is part of the problem.

Well, I’m sorry that what looked like a very clear and concise OP to me has been next to incomprehensible to at least two people.

I do understand that government spending tax money means that GDP is at least kept constant. I understood that before I even wrote the OP. What you point out about GDP likely increasing with more government spending relative to less government spending, because people tend to save more than the government does, actually goes along with the point I was trying to make.

@Neodoxy

The significance of Bastiat’s fallacy is twofold:

1. It shows that even if people aren’t forced to spend their money on one thing that they are likely to spend at least part of it, and therefore nothing would be waisted (I’d argue that part of Bastiat’s failure is that he didn’t properly take into account the consequences of savings).

2. That simply spending does not mean an increased in happiness. Even if GDP did decrease as a result of the window not being replaced then wouldn’t people still be better off if they only spent half the money on other things? Of course, a large part of this depends upon price flexibility and the multiplier.

People (including, it seems, econometricians) equate increased GDP with increased economic growth, increased wealth, increased happiness, etc. My point was to (try to) help show how that equation is misleading.

[Aristippus]

I don’t see what’s wrong about what Autokylos is saying. The point is that in both scenarios you have the same GDP, but in one scenario - the one in which the window is not broken - you have a greater amount of goods and therefore a greater satisfaction of wants. Therefore GDP and GDP growth is misleading if it is taken to actually measure well-being.

[Wheylous]

People (including, it seems, econometricians) equate increased GDP with increased economic growth, increased wealth, increased happiness, etc. My point was to (try to) help show how that equation is misleading.

With this I agree. It did not appear that this is what you meant, however. Especially not with this:

Now let’s change the scenario to something a bit more modern. Let’s say that the government of Parableville takes six francs from James Goodfellow in taxes. It then spends that six francs on a new window from the glazier. Following the definition of GDP, this would raise Parableville’s GDP from 100 francs to 106 francs (ceteris parabus). Once again, however, this doesn’t take into account the loss against James Goodfellow, which is also six francs (the amount he was taxed). Offsetting this loss, and holding to ceteris paribus, brings the GDP back down to 100 francs.

What it seems you are trying to say is that buying the window increases GDP to 106 but the money transfer away from James brings it back down, which is not the case. The window was still produced. Even if it was useless.

The point is that in both scenarios you have the same GDP, but in one scenario - the one in which the window is not broken - you have a greater amount of goods and therefore a greater satisfaction of wants. Therefore GDP and GDP growth is misleading if it is taken to actually measure well-being.

​Yes, this is what the original fable discusses - Total wealth. I agree. But it’s not GDP. I agree that GDP as a measure of stuff produced doesn’t necessarily mean wants are satisfied as much as possible (because, by its very nature, government spending doesn’t give a choice except through democracy, which is both inefficient and assumes some sort of “average utility-maximizing policy”)

[Aristippus]

As it seems to me, Autolykos was measuring the GDP of an hypothetical broken window scenario in order to explain why GDP is misleading.

[Neodoxy]

Autokylos, I’m afraid that I misread and misinterpreted your post. I thought that you were arguing that the taxation transfer would lead to an explicit, rather than an implicit loss of GDP. Upon first reading the post it seemed like you were presenting an argument against the specific methods used in GDP rather than the very concept of GDP (arguably you were doing the former but you were attacking a bedrock concept of GDP rather than simply one error in accounting).

"Well, I’m sorry that what looked like a very clear and concise OP to me has been next to incomprehensible to at least two people.

I do understand that government spending tax money means that GDP is at least kept constant. I understood that before I even wrote the OP. What you point out about GDP likely increasing with more government spending relative to less government spending, because people tend to save more than the government does, actually goes along with the point I was trying to make."

What is the point of this part of the post?

[Autolykos]

@Wheylous

With this I agree. It did not appear that this is what you meant, however. Especially not with this:

Now let’s change the scenario to something a bit more modern. Let’s say that the government of Parableville takes six francs from James Goodfellow in taxes. It then spends that six francs on a new window from the glazier. Following the definition of GDP, this would raise Parableville’s GDP from 100 francs to 106 francs (ceteris parabus). Once again, however, this doesn’t take into account the loss against James Goodfellow, which is also six francs (the amount he was taxed). Offsetting this loss, and holding to ceteris paribus, brings the GDP back down to 100 francs.

What it seems you are trying to say is that buying the window increases GDP to 106 but the money transfer away from James brings it back down, which is not the case. The window was still produced. Even if it was useless.

You’re absolutely right that the standard definition of GDP only accounts for production. I guess I should’ve been clearer, sorry. What I meant was that if GDP took into account economic losses as well, its results would be different.

@Wheylous

Yes, this is what the original fable discusses - Total wealth. I agree. But it’s not GDP. I agree that GDP as a measure of stuff produced doesn’t necessarily mean wants are satisfied as much as possible (because, by its very nature, government spending doesn’t give a choice except through democracy, which is both inefficient and assumes some sort of “average utility-maximizing policy”)

Right, GDP doesn’t actually measure total wealth, satisfaction of wants, etc. It just measures what’s produced. It pays no attention to whether anything produced is produced because of a genuine increase in valuable output, or if it was produced due to a “broken window”.

[Autolykos]

@Neodoxy

Autokylos, I’m afraid that I misread and misinterpreted your post. I thought that you were arguing that the taxation transfer would lead to an explicit, rather than an implicit loss of GDP. Upon first reading the post it seemed like you were presenting an argument against the specific methods used in GDP rather than the very concept of GDP (arguably you were doing the former but you were attacking a bedrock concept of GDP rather than simply one error in accounting).

No problem. I’m sorry I wasn’t clear enough in the OP. It seemed crystal clear to me, but I’m hardly everyone.

My point was to change the definition of GDP to illustrate how it doesn’t account for “what is unseen”. Basically the intended format was this: “GDP, as actually defined, shows us one result, but if it was defined this other way, it would show us a different but actually more accurate result”. In other words, I wanted to show how the actual GDP measurement employs the same kind of economic fallacy as that elucidated by Bastiat in his parable of the broken window. Does that make sense?

@Neodoxy

"Well, I’m sorry that what looked like a very clear and concise OP to me has been next to incomprehensible to at least two people.

I do understand that government spending tax money means that GDP is at least kept constant. I understood that before I even wrote the OP. What you point out about GDP likely increasing with more government spending relative to less government spending, because people tend to save more than the government does, actually goes along with the point I was trying to make."

What is the point of this part of the post?

Sorry if that sounded harsh or snarky (or both) - that wasn’t my intention. What I meant was that I genuinely was sorry that I failed in clearly articulating my thoughts in the OP. I then tried to rectify that failure by trying to dispel any notions that I misunderstand what GDP is all about.