The city as a glorious market failure

Liked it? Spread the word...

“The city is an economic engine”. This built-in assumption is now rooted in so many studies that it had become taken for granted. To the point where no one pauses to ask whether or not it’s true anymore. If you are as fascinated with cities as I am, it is clear to you that this claim is pivotal, so why are its supporting arguments so vague and easy to refute?

Quite a few studies demonstrate that large cities are associated with high levels of productivity. This statistical correlation is the basis for the assumption that the city is an economic engine. This is the claim that I intend to dismantle. I am not arguing with the facts, only with the conclusions made by the various researchers.

To facilitate the discussion I’ll present the data regarding the conventional position from the following famous article. In my opinion, it is a must-read article for anyone in the field. It’s nothing short of brilliant.

Growth, innovation, scaling and the pace of life in cities
Luís MA Bettencourt, José Lobo, Dirk Helbing, Christian Kühnert, and Geoffrey B. West
PNAS April 24, 2007. 104 (17) 7301-7306

Quoting the main paragraph:
“The most striking feature of the data is perhaps the many urban indicators that scale superlinearly. These indicators reflect unique social characteristics with no equivalent in biology and are the quantitative expression that knowledge spillovers drive growth, that such spillovers in turn drive urban agglomeration, and that larger cities are associated with higher levels of productivity. Wages, income, growth domestic product, bank deposits, as well as rates of invention, measured by new patents and employment in creative sectors all scale superlinearly with city size, over different years and nations with exponents that, although differing in detail, are statistically consistent. Costs, such as housing, similarly scale superlinearly, approximately mirroring increases in average wealth.”

… and in plain English

1. Bigger cities are more efficient
OK, we can understand that claim. It is the economies of scale that can be demonstrated in industrial plants. The article compares it to mammals. The elephant’s energy efficiency is better than that of a mouse, although in principle both have the same biology of blood vessels, heart, lung, and so on.

2. Slightly bigger cities are Way More efficient
This is the translation of the “super-linear” above.
It means that once the city starts to grow, it will go on and on and grow infinity as the bigger it gets the growth engine is only speeding up, pressuring for even more growth. The authors of the article write that this is a process that must collapse. There is no parallel for it in mammals. It is not sustainable.
No Urban planner can read this claim and stay calm. We must understand this mechanism.

3. Slightly bigger cities are way more efficient because of knowledge spillovers
Ding! Not actually proven.. or rather doesn’t make sense.

Let’s explain again from the start

The claim is that If we take a million residents and scatter them in five cities, they will produce X. On the other hand, if we put these million people and settle them in one city, they will produce more! 10 to 30 percent more!

Note 1: No matter what index we use, whether it’s paycheck, the number of inventions, the number of bank deposits … All of these metrics behave the same way.

Note 2: The data is consistent throughout history and in different cultures.

Note 3: The data is consistent even when the size of the cities was 100,000 inhabitants and we compare them to the villages around it, and when cities are monsters of 40 million and we compare them to the “small” cities of five million inhabitants.

And the question arises from the study:
What the Fuck?!?

Why the hell are a million people in a crowded group producing more than a million people scattered around? Do they Invent more? Get paid more?

Knowledge spillover

The researchers’ answer is based on a phenomenon called “knowledge spillover” (knowledge leakage). This phenomenon refers to the knowledge transferred between companies, organizations and people in an informal way. For example, a marketing manager of one company quits and goes to work for the competitor, doctors from different research centers meet and talk at the same conference, etc. In these cases knowledge is being transferred and can increase creativity and facilitate development. Knowledge spillovers can also occur between fields, for example between a meteorologist and an economist where a method for predicting ​​the weather ignites an idea regarding stock market behavior.

The claim is that knowledge spillovers are more prevalent in larger cities. In other words, the larger the city, the more knowledge flows it will have and therefore allow for better creativity, more output and higher wages.

Have you noticed the logical creak?

I mean … it makes sense until you stop for a moment and think about it.

If the claim about knowledge spillovers was true, the advantage should be for denser cities, not larger. This is not the case.

If the claim was true, the advantage it demonstrates should have weakened at a certain size. After all, a person does not really meet more people in a city of 20 million than in a city of 5 million.

If the claim was true, in the present era, when the exchange of information is so quick and so easy, when knowledge is carried out regardless of physical location, the power of the cities was supposed to weaken and not become stronger. This is not what happens.

In the language of the researchers, it can not explain a “super-linear” growth.

The conclusion is that knowledge spillovers can’t be the mechanism behind “super-linear” growth. I claim that we confuse statistical correlation with causality. It’s like saying that kids with an iPhone do better at school. Yes, that’s true. Statistically, if we’ll check, we’ll find that children with an iPhone do better than children who do not have the device.
But it does not (!) mean that the iPhone is the reason for academic success. It is much more likely that children with iPhones come from a higher socioeconomic background and the correlation between parental success and children’s success has long been proven.

Which of the two claims do you find to be more logical?

A. The fact that the city is crowded is the reason for the success of businesses in it.


B. The success of businesses in the city is the reason for people to flock to it and make it crowded.

I am not arguing with the facts, only with the conclusions of the various researchers. Yes, the salaries in the big city are higher than in the periphery. The measured output in the city is greater, the “everything” is greater. But it is not due to density. Density itself has no economic advantage. Crowding is a side effect created in the city because of the success of its businesses.

A dumb example

Million people eat a million hamburgers every day and only at McDonald’s. Half of the people live in the periphery and half in the city. All McDonald’s restaurants are located in the city and people come to eat in them both from the city and the periphery.

We conclude that:

  • People in the city eat two meals a day and people in the periphery do not eat at all.
  • The city center is the reason that the production of hamburgers is efficient.
  • If the periphery was as effective as the city we would produce 2 million hamburgers a day.

But that’s all nonsense.
The only reason all the restaurants are concentrated in the city center is that people from the periphery are willing to go downtown to eat. If they were not willing, McDonalds would have opened a restaurant in the periphery and would produce there the same hamburgers at the same price and same efficiency. A million people eat a million burgers regardless of the location of the restaurants. The city never created higher demand nor higher supply in the overall market.

Some real life examples

Detroit was crowded and crashed, Palo Alto in the silicon valley is scattered and became successful, Albuquerque is a huge city with very low density, and on and on. you don’t need me to find more evidence for it in your country.

What is the reason for the seeming success of businesses in the city?

A city is essentially a market (of products, services, opinions). The city is a physical manifestation of a market failure in which it is more profitable for the supply of goods to crowd without actually creating value for the entire market.A million residents will produce and consume the same X if we’ll spread them in five cities or settle them in one. Our illusion is created when we look at the city versus the periphery. In other words, if we take half a million residents and put them in the city and scattered the other half in it’s periphery, we will find very quickly that even though the people are scattered, the businesses are crowded in the city. When we’ll measure the city and compare it to the suburbs it will seem to us that the city is more efficient. It is not, it simply attracts economic activity to it and depresses economic activity around it. The overall remains the same X.

So why should businesses be skewed to the center of the market if they gains nothing?

In a nutshell, they do not gain productivity, they avoid loss of market share. This market failure is presented in more detail in the mathematical model I have shown here. It is the extension of the “Ice Cream Vendors” model and presents Nash equilibrium in a space with multiple sellers and multiple consumers.

Liked it? Spread the word...

Leave a Reply

Your email address will not be published. Required fields are marked *