The law of diminishing returns states that adding more of one input to a production process, while keeping everything else the same, will eventually produce smaller and smaller gains in output. It’s one of the most fundamental ideas in economics, and it shows up everywhere from farming to fitness to software development.
How It Works
Imagine a small kitchen with one oven. Hiring a first cook dramatically increases how many meals get made. A second cook helps too, since they can prep while the first one plates. A third cook adds a bit more output. But by the time you’ve crammed a sixth or seventh cook into that same kitchen, they’re bumping into each other, waiting for oven space, and barely moving the needle on total meals served. Each additional cook contributes less than the one before.
That’s the core mechanism: one input (labor) keeps increasing while another input (kitchen space and equipment) stays fixed. The gains from each new unit of the variable input shrink over time. Economists call the extra output from each additional unit the “marginal product,” and the law says that marginal product eventually declines.
This is a short-run concept. It specifically applies when at least one factor of production is held constant. The kitchen has a fixed number of ovens. The factory has a fixed amount of floor space. The farm has a fixed number of acres. In the long run, when you can change everything at once (build a bigger kitchen, expand the factory), different rules apply.
A Classic Example: Fertilizer and Crops
Agriculture is where economists first noticed this pattern. Early thinkers like Turgot, Malthus, and David Ricardo all observed that working the same plot of land harder and harder didn’t keep producing proportional results. The soil, the sunlight, and the water set natural limits.
Modern data confirms this clearly. In China, every kilogram of fertilizer (measured as pure nutrient) increases grain production by about 7.5 kilograms on average. But that average masks a curve. At low fertilizer levels, each additional kilo makes a big difference. As application rates climb, the gains shrink. Research on Chinese grain-producing regions found that in areas already using optimal fertilizer levels, increasing application further provided no additional yield benefit at all. In fact, the promoting effect of fertilizer on grain yield dropped by 0.222 units for every one-unit increase in the proportion of grain-producing area, a clear sign that those regions had hit the wall of diminishing returns.
Software Teams and Communication Overhead
One of the most famous illustrations outside economics comes from software engineering. Brooks’s Law states that adding more people to a late software project actually makes it later. This sounds paradoxical, but the mechanisms are pure diminishing returns.
New developers need time to learn the codebase, during which they’re unproductive and pulling experienced team members away to answer questions. As the team grows, communication overhead increases dramatically. A team of 5 has 10 possible one-on-one communication channels. A team of 15 has 105. All that coordination time comes directly out of time that could be spent writing code. Tasks also get fragmented into smaller and smaller pieces to distribute among more people, which introduces bugs at the seams and slows integration. Past a certain team size, each new hire subtracts more productivity through coordination costs than they add through coding.
Diminishing Returns in the Gym
If you’ve ever wondered why your first few months of working out produced visible results and then progress seemed to crawl, diminishing returns explains it. Research confirms a dose-response relationship between training volume and muscle growth, but the curve flattens as volume climbs.
A study published in Medicine and Science in Sports and Exercise split trained men into three groups: one set per exercise per session, three sets, or five sets. For upper arms, five sets per session produced significantly more growth than one set. But the jump from three sets to five sets showed only weak additional benefit. For the thighs, five sets again beat one set convincingly, and there was moderate evidence favoring five sets over three. The pattern is consistent: tripling your volume from low to moderate helps a lot, but going from moderate to high helps less.
A meta-analysis found a clear dose-response curve up to about 10 weekly sets per muscle group but couldn’t determine what happened beyond that threshold due to limited research. The practical takeaway is that your first several sets each week do the heavy lifting for muscle growth. Additional sets still contribute, but each one adds a smaller slice of progress, and at some point the recovery cost outweighs the growth stimulus.
How It Differs From Economies of Scale
People often confuse diminishing returns with the idea that bigger operations become less efficient. These are related but distinct concepts operating on different timelines.
Diminishing returns is a short-run phenomenon. You’re changing one input while others stay fixed. A factory adds night-shift workers but doesn’t buy more machines. A farmer applies more fertilizer to the same field. The constraint is that something remains constant.
Economies of scale (or diseconomies of scale) describe what happens in the long run, when all inputs change together. A company builds a second factory, doubles its workforce, and upgrades its supply chain simultaneously. Output might grow proportionally, more than proportionally, or less than proportionally relative to the increase in inputs. That long-run question is about returns to scale, not diminishing returns in the classical sense.
The distinction matters because diminishing returns are essentially inevitable in the short run. You will always hit a point where cramming more of one input into a fixed system yields less. Returns to scale, on the other hand, can go in any direction depending on the industry and how well the expansion is managed.
Digital Products and Near-Zero Marginal Cost
Digital goods seem like they should break the law entirely. Once an app or ebook is created, the cost of distributing one more copy is essentially zero. There’s no field to over-fertilize, no kitchen to overcrowd. So where do diminishing returns fit?
The key is that digital products are characterized by high fixed costs and near-zero variable costs. Someone still has to build the app, write the book, and maintain the servers. The production side still faces diminishing returns in familiar ways: adding more programmers to a software project hits Brooks’s Law, and adding more writers to a single book doesn’t make it twice as good. What changes is the distribution side, where marginal costs are so low that the usual supply-side constraints barely apply.
This doesn’t eliminate scarcity so much as shift it. Digital products compete fiercely on price precisely because reproduction is cheap, which means the returns to marketing, user acquisition, and platform fees all follow their own diminishing curves. Google’s chief economist Hal Varian has argued that the digital economy doesn’t require entirely new models. The same principles apply, just with the fixed and variable cost balance tipped heavily toward fixed costs.
Why It Matters for Everyday Decisions
The law of diminishing returns is ultimately about optimization. It tells you that there’s a sweet spot for any input, and pushing past it wastes resources. The first hour of studying for an exam improves your score more than the eighth hour. The first coat of paint on a wall does more than the fourth. Your first cup of coffee wakes you up more than your third.
In business, recognizing the point of diminishing returns helps companies decide when to stop hiring for a project and start investing in better tools instead, or when to stop spending on advertising and redirect budget elsewhere. In personal life, it helps explain why balance tends to outperform obsessive focus on any single input. The principle doesn’t say “stop adding input.” It says “know that each additional unit buys you less, and plan accordingly.”