The Error of Aggregates - Island of Sanity

Island of Sanity

Policy Issues

The Error of Aggregates

The concept

Two of my economics textbooks in college talked about the "error of aggregates". The idea is that something could be good if done by one person or a small number of people, but become bad if done by many people. That is, you cannot necessarily "aggregate" -- add up -- the pros and cons of any given action. If one hundred people do something, the result is not necessarily one hundred times the effect if one person did it.

(I've never seen this idea mentioned anywhere else. I just did a Bing search and couldn't find any hits that were talking about the same concept. But it showed up in two textbooks by different authors, so it wasn't just one guy's idea. Whatever.)

I suppose there's a sense in which this is obviously true. If one person in a country makes bicycles, that's a good thing, at least for people who want bicycles. But if everyone in the country made bicycles, then who would be producing books and houses and food? We can't all do the same thing. But the problem there is not really that it's bad to do X, but that's it's bad to fail to do everything else. That's not the point of the "error of aggregates".

Trashing the highway

One of the books gave this hypothetical example to introduce the idea:

Suppose that instead of paying for garbage collection, someone just throws his trash out his car window onto the highway. The environmental impact of one person doing this, the author said, would be negligible. But of course if everyone did it, our highways would be a mess, causing pollution and disease. Thus, he said, if done by just one person it would be good, but if done by many people it would be bad.

I don't see how this analysis fits the scenario. This person saves some money by imposing a cost on others without their consent. As he's throwing his trash on a public highway, presumably the government will end up paying to clean it up, with the cost passed on to the taxpayers. Suppose that instead of throwing the trash on the highway he threw it in his neighbor's yard. Then -- assuming that for some reason the neighbor can't call the police but has to deal with the trash -- he is forcing his neighbor to pay for his trash pickup. By saying he throws it on the highway he is just dividing the cost among many taxpayers instead of a single victim, but the principle is the same.

The only "good" here is that an inconsiderate person manages to force others to pay for something that he should be paying for himself. Actions like this are generally considered crimes.

There is no "error of aggregates" here. If one person did this, he would save a small amount of money while imposing a small cost on his neighbors. If many people did this, they would save a large amount of money while imposing a large cost on their neighbors. It "aggregates" exactly how you would expect: Two people doing this would do twice as much harm as one person doing it, etc.

Trashing savings

In both books, the point they were building up to was this:

If one person saves his money, it is good. He prepares to send his children to college or plans for retirement or whatever. But if many people start saving their money, it is bad. Reducing total consumption means that stores and manufacturers don't do as much business, people lose their jobs, and the economy falls into recession.

But this analysis only sounds plausible as long as you think of only the benefits of one person saving and only the downside of many people saving.

At the individual level, if ten people save for retirement, they each get the same benefit that one person gets by saving for retirement. So the benefit to ten is ten times the benefit to one. But okay, that's not the point. They're talking about the affect on the overall economy.

What happens to one person's personal financial condition when he increases his saving? He gains the benefit that this money will be available at a future date when he needs it. If he invests the money wisely, he'll get his money back with interest or profits. But he pays a price for this: he loses the use of his money now. When you save, you might give up buying a new car today so you can send your kid to college in ten years. Yes, your kid gets to go to college. But you don't get the new car. Or if you just save money as an investment, you give up, say, $1000 today so you can have $2000 in ten years.

At the macroeconomic level, if many people increase their savings, what happens? They spend less on consumer goods. Stores and factories that deal in consumer goods lose money. They may have to lay off employees, or even go out of business. But where does the money go? Does it evaporate into the air? Of course not. They may invest it, for example in the stock market. Or they may put it in the bank. The bank then loans the money out, mostly to businesses. In either case, it ends up being invested in businesses. They then use this money to build new factories, buy equipment, engage in research and development of new products, etc.

It is not at all clear that increased savings will cause a recession. When people spend now rather than save, they buy consumer goods, like cars and furniture and books and candy. When people save rather than spend, the money is invested in businesses, which buy capital goods, like factories and laboratories and office equipment. Either way the money is still spent. People are still employed.

Thus, there is a close parallel between the effect of higher savings on a personal level and on a macroeconomic level. On a personal level, a person gives up immediate spending for investment in a future benefit. On a macroeconomic level, we give up immediate consumer spending for investment in future economic growth.


Maybe there are real, non-trivial cases of the error of aggregates. I can't think of one. Saving money isn't one.

© 2014 by Jay Johansen


Sasquatch May 23, 2014

You may have to look deeper than just the financial tradeoff to see an error of aggregates. Think about the diversion of resources caused by many people doing one thing. Pretty much any fad will draw away resources that could have gone to other things or that could have been left for future uses. Since fads tend to be short-lived, much of what is produced to sustain them will wind up as waste as people move on to the next fad. So we end up with millions of hula hoops rotting in landfills-- untold gallons of petroleum products rendered unavailable for more important purposes. So while the hula hoop companies made money and created jobs, they also drove up the cost of the raw materials for syringe makers. A few hula hoopers wouldn't have made a dent in the cost of plastic. But millions?

Dale Dec 30, 2014

What you say is true, but not the point. The idea of the "error of aggregates" is that something is good for one person to do but becomes bad if many do it. But spending on a fad, like hula hoops, to the extent that it wastes resources, wastes a little bit of resources if one person does it and wastes a lot if lots of people do it. It's not good becomes bad, but a little bad becomes a lot of bad. (You could say that resources used on a fad like this are not "wasted" if people got pleasure out of it. Isn't the whole point of an economy supposed to be to enhance human happiness? But to the extent that it's a waste, it's a waste big or small.)

Add Comment