Economies of Scale

Big things work better than small, but only sometimes

Economies of scale is a term used to describe situations where doing more of something make the per unit cost go down. This is generally due to the increased volume making specialized capital more cost effective, but also sometimes reductions in overhead or bargaining costs. This is the principle behind specialization leading to massive increases in output.

For example, if I make just myself a ham sandwich every day for lunch I will probably just use a knife to cut the ham, or spend the extra money to buy it pre-sliced. I either spend more time cutting or more money to avoid cutting. If I am going to make ham sandwiches for everyone in the office everyday, however, I will make or buy a meat slicing machine, allowing me to save money on pre-sliced ham and time on doing it by hand with a knife. The machine costs money, perhaps as much as 10,000 sandwiches worth of pre-sliced ham. That would take me over 27 years of sandwiches for myself to pay off, but if I am making 100 sandwiches a day it pays for itself in about 3 months, and that is before I consider the value of the time it saves. So, if all 100 sandwich eaters makes their own sandwich we can’t benefit from the scale, but if I specialize and make all the sandwiches we can make them faster and cheaper.

Likewise, government regulations often create economies of scale in businesses. Imagine a new rule requires businesses to have a compliance officer to make sure it is being followed, and how much of his time it takes scales with the size of the firm. A firm with 50 employees takes 1 full time compliance officer, and a firm with 25 only takes a half time compliance officer. Assuming there are zero other scale effects, firms with <50 employees are going to be at a disadvantage. Whereas before two 25 employee firms were just as efficient as one 50 employee firm, now each 25 employee firm needs to hire a compliance officer whereas the 50 employee firm does the same. In other words, it takes two compliance officers to keep two smaller firms in compliance and only one for a larger firm. Unless the smaller firms can pay their officers less or do some part time consulting thing, we would expect them to merge or go out of business competing with the more efficient larger firm.

So, economies of scale come from both capital efficiency and oversight efficiency. Importantly, however, we need to note that they are not always increasing! Bigger might be always better along some margin, but it is often negative along others. In fact there is lots of evidence that scale breaks down very quickly. There are many more types of small animal than large. Ants are extremely successful at their size, but dog sized ants don’t work, both for biological engineering reasons (insects need a LOT of oxygen to get big due to inefficient respiratory/circulatory systems) and biological nice reasons (it would take a lot of food to feed 10,000 dog sized ants.)

Likewise, you will likely never come across a hair salon the size of a Walmart. You could imagine one, a giant cavernous salon with rows of hundreds of chairs, the snip of scissors and the buzz of clippers nearly drowning out the awkward small talk, the stink of floral shampooing and talcum powder nearly overwhelming. You will probably never see one, however, because hair cutting doesn’t scale well. Sure, having two or three hair stylists sharing a space makes sense: you can share a cash register and rent, and walks ins don’t want to wait too long so if you can’t cut their hair there is no loss if someone else does and puts some of that money towards your rent, volume discounts on talc, someone to chat with during a slow patch etc. However, no matter how much capital equipment you invest in it is difficult to increase the speed of haircutting or reduce its cost; really nice scissors and clippers help a little, but then that is about it. Plus, customers like things such as getting the same person they trust, being in a smaller more peaceful environment, and not having to drive too far. All those cut against having a giant salon and suggest that many smaller scattered salons work better.

Even industrial businesses struggle with this issue. Sure, it is easier to make a ton of steel when you are making 1,000 tons at a time than 100 pounds, and more than 100 times easier, the capital aspect takes care of that. The employee management side pushes back, however.

This is referred to as the “principle agent problem”, which is just a fancy way of saying that getting people to do what you want is hard, even when you pay them, because they want different things. Well, maybe not fancy but more concise way of saying it. If I hire someone to do a job for me, I probably can’t just walk away and expect the work will be done 100% perfectly. I have to spend some time supervising to make sure they are doing the job, answering questions, redirecting them if they are doing it wrong, etc. This situation has economies of scale at a very low level, but they disappear quickly.

Imagine that we here at Hammer Scientific Industries consist entirely of me. (My pronouns are we/our.) I make wargaming miniatures (man dollies) and due to extremely high demand need someone to help mix and pour the resin, demold the bits, make molds, etc. I hire Steve, and can supervise him at basically zero cost because he is standing right next to me as we work. The only time he is out of my sight is when we use the bathroom, and if we time it right even that isn’t a problem for long. Production nearly doubles, as pretty quickly Steve is about as good at the job as I am.

This is clearly a wonderful solution, suggesting I need to hire more! I enlist Kyle and Pierce in my ever growing empire of men making tiny men for other men. I can still keep an eye on them easily because we are all in the same place working elbow to elbow, but it is a bit harder since concentrating on my work and the work of three other people is a bit tough. More questions, more distractions, and because all of us have elbows and such, we need to spread out the workspace. When I need to do businessy things like accounting and order fulfillment I have a lot of trouble just plopping a laptop down on the work bench, and so retreat to an office, and who knows what trouble those buggers get up to then. Still, I am at nearly 4x the production I was at before! Hooray!

The orders though… the orders keep stacking up! Damn my success! I could sell three times what I am doing now! I hire another seven people, Maria, Sebastian, Olivia, James, Madeline, Evelyn, and Millicent, and set them to work. Work space was an issue, but I rented the space next to mine and so that scaled nicely. What is a little less nice is that I am having a really hard time keeping and eye on everyone, helping them out and solving problems, and doing anything else myself. I mean, it doesn’t help that all the new employees are under the age of 10, but even if they were all adults if each of 11 employees need to ask me a question that results in a three minute discussion every hour, that is over 50% of my time just on questions, much less coordination and making sure they are doing things right. I mean, Millicent is a good worker and means well, but damn if she doesn’t have a bossy streak that sets the others off on wrong paths if you don’t correct constantly. Arguments are rare, but after the juice box incident last week, things are getting crazy for me. I am lucky if with 10 employees I can get 8 times the output of just myself.

Ahh, but this is a promising idea… what if I promote Steve to be a “senior production lead”? Sure, he still has all his old responsibilities, but I can whack on “Watching everyone else while I try and get some other work done” to his list. In essence he can share the organizational role I am doing, getting paid an extra nickel an hour. Brilliant! For this small cost I can get reliably to 9 times the output of just myself.

But people keep buying more! GAH! I need to hire more people still! Now, Steve and I could supervise 5 employees each and still get stuff done ourselves, so if I hire 25 total workers I need 5 managers… and I can’t be one of them because I need to manage the managers AND keep an eye on everything else. So with 25 line workers, 5 managers and me I can probably make at best ~25-30 times what I could make all by myself, but more likely ~20 times. Why just 20? Well, obviously I can’t expect Steve, Kyle or Pierce to be as amazing at supervising as I am, and while I promoted Olivia she might be being taken advantage of by her underlings. The new workers, well there are so many it is getting hard to even remember all their names much less keep track of their work habits and judge whether their managers are doing a good job managing them. I don’t even know if half the employees want to keep making man dollies, or think they really work doing something else. If mistakes are made it is increasingly difficult to know who made them, and the managers seem to blame each other as much as their own teams when things go wrong… man it was so much nicer when it was just me, Steve, Kyle and Pierce…

Capital technology, in the form of machines, infrastructure and specialized job roles all provide economies of scale. Organizational problems such as principle/agent issues tends to push back the other direction. There is also the third issue of what it is the organization even looks like. Large groups of people are going to include large numbers of differing preferences. Some people like working for large corporations, some prefer small, and some hate having a direct boss at all and want to run their own company. Some people like big company picnics and mandatory fun, others want to come to work and go home, never mixing the two. How many people want to live in the small town you founded your business in? If that isn’t enough to staff everyone, you might have to move HQ to a larger city to find enough talent. The list goes on.

All told, economies of scale are real and important, and so are diseconomies of scale. We often get excited for bigger more efficient businesses, schools, governments what have you, forgetting that the efficiencies happen on certain margins while inefficiencies happen on others, and we always have to trade them off against each other.