# Economics #4: Economies of Scale

What are some industries are characterized by many small firms in fierce competition (retail, restaurants, etc) while some industries are characterized by a few large firms with market power (telecommunications, railroads, etc)?

What is it about the restaurant industry that allows competitors to open up a shop across the street. What is it about the telecommunications industry that often limits competition to one or two providers in a city.

That answer these questions we will start with the concept of economies of scale and then apply that to industrial organization.

### Returns to Scale

To understand economies of scale on the cost side, we first need to understand returns to scale on the production side. Imagine a fictional firm that produces unicorn robots. Let’s say that the only input required for producing a unicorn robot is a human employee. We start with one employee producing one robot.

Now let’s see what happens if we double our inputs from one employee to two employees. Either our output will more than double, double, or less than double.

If our output more than doubles, we are experiencing increasing returns to scale. Getting bigger is better. If our output doubles, we are experiencing constant returns to scale. Getting bigger doesn’t help. If our output less than doubles, we are experiencing decreasing returns to scale. Getting bigger hurts.

So if our unicorn robot firm adds an employee, maybe the two can both specialize in one part of the process and increase the rate at which they finish whole robots. Instead of each employee producing one unicorn robot on their own, maybe they produce three all together when they work together.

### Economies of Scale

Now for the cost side of things. Imagine there are two of those unicorn robot firms. One produces unicorn robots the way we described earlier, and one produces unicorn robots with a giant automated factory.

The factory model costs millions of dollars to get started while the employee based model only costs a few thousand dollars. Let’s give this some solid numbers.

Firm Total Cost Output Average Cost
Small Firm \$1,000 10 \$100
Large Firm \$1,000,000 20,000 \$50

Even though the total cost of the large firm is higher, their average cost per unicorn robot is lower. This is the idea behind economies of scale. The long-run average costs fall as more output is produced. If the large firm wanted to, they could undercut the selling price of the small firm and take their market share. The small firm would have to decide whether to invest in scaling up or to go out of business.

If getting bigger is also cheaper, why then doesn’t one giant company just dominate the world? Companies face a similar pattern of long-run average costs where they start out experiencing economies of scale, then hit a size where they experience constant returns to scale, then if they continue to grow, hit a size where they experience diseconomies of scale.

When a firm gets too big, it becomes increasingly difficult to manage. Complexity and bureaucratic inefficiencies cost money. So how big should a firm get? Well based on what we talked about so far it should just keep producing until its long-run average costs are minimized (minimum efficient scale). But this is missing one key factor – demand.

### Industrial Organization

Consider a local pizza restaurant. What is stopping them from hiring a bunch of employees and buying a bunch of high-tech machines in order to produce millions of pizzas? Well, there is no demand for a million pizzas.

There are three main determinants of an industries structure: minimum efficient scale, demand, and barriers to entry. Here we are only looking at the first two.

The minimum efficient scale is just what we were talking about earlier. It is the minimum level of production where a firm can minimize long-run average costs. In the above picture, that occurs at Q2.

What happens if demand in the industry is greater than Q2? For example, what happens if the minimum efficient scale for any one firm in our unicorn robot industry is 10 but the market demands 1000? Well, we would have 100 firms each producing 10 unicorn robots. This is simplified, but the concept holds true. Consider industries like food service, where the quantity produced to hit the minimum efficient scale is low and the demand is high. We get a restaurant on every corner.

What happens if demand in the industry is lower than or equal to Q2? For example, what happens if the minimum efficient scale for any one firm in our unicorn robot industry is 100 but the market only demands 90? Well, we would have 1 firm producing 90 unicorn robots. Consider industries like telecommunications where the the quantity produced to hit the minimum efficient scale is high and the demand is about even with that. We get one telecom provider per city (natural monopoly).

Another way of thinking about this is that industries with high fixed costs (property, plant, equipment, etc) will have a higher minimum efficient scale to cover those expenses. Industries with low fixed costs (lease, no equipment, etc) will have a lower minimum efficient scale.

### Conclusion

The ideas of economies of scale, minimum efficient scale, and demand give us a framework for understanding how industries will organize themselves. If the efficient level of production for a single firm is less than the market demand, we will get an industry with a lot of firms. If the efficient level of production for a single firm is more than or equal to the market demand, we will get an industry with only a few firms that dominate.