The crowded kitchen
In the last guide we watched a firm turn inputs into output through its production function. Now hold one input fixed and pour in more of another, and something almost inevitable happens. Picture a small pizzeria with one oven, one bench, one set of pans — these are the fixed factors of production for tonight. The first cook you hire is wonderful: she does everything and turns out 10 pizzas an hour. Add a second and they split the work — prepping while the other bakes — and output leaps to 24. Add a third and you reach 33. So far each new pair of hands is a gift.
But the oven still bakes only so many pizzas at once. The fourth cook lifts you from 33 to 40 — a smaller gain. The fifth gets you to 45, the sixth to 48, and by the seventh you are at 49: cooks now jostle for the one oven, trip over each other, wait their turn. Notice nobody got lazy and the kitchen did not break. The *extra* output from each new cook simply shrank, because every cook had less and less of the fixed oven to work with. That shrinking extra is the heart of today's idea.
Three views of the same numbers
Economists watch this process through three measures, and the trick is that all three come from a single column of numbers. Total product is just the total output at each number of cooks — our running tally of 10, 24, 33, 40, 45, 48, 49. Marginal product is the *extra* output one more cook adds: the gap between one row and the next. And average product is total product divided by the number of cooks — output per head. Keep the names straight and the rest is bookkeeping.
Cooks Total product Marginal product Average product 1 10 10 10.0 2 24 14 12.0 3 33 9 11.0 4 40 7 10.0 5 45 5 9.0 6 48 3 8.0 7 49 1 7.0 Marginal product = (this total) - (previous total) Average product = (this total) / (number of cooks)
Read the marginal column on its own and the law jumps out: 10, 14, then 9, 7, 5, 3, 1. It rises at first — the second cook fits the fixed kitchen beautifully — then turns and falls, step after step. The point where marginal product stops climbing and begins to drop is precisely the onset of diminishing marginal returns. Crucially it set in at the *third* cook, while output was still healthy and rising. Diminishing returns is not a wall you hit; it is a slope that gets gentler with every step.
How marginal pulls average around
There is a tidy relationship between the marginal and the average that is worth holding onto, because it shows up everywhere — in test scores, in cricket batting, and very soon in cost curves. The rule: whenever the marginal is *above* the average, it pulls the average up; whenever it falls *below*, it drags the average down. Think of your grade average. Score above your current average on the next test and your average rises; score below it and your average sinks. The new test is the "marginal"; your standing is the "average."
Look back at our table with that lens. At the second cook marginal product (14) sits above average product (12.0), so the average climbs from 10.0 to 12.0. By the fifth cook marginal (5) has fallen below average (9.0), so the average is now sliding. It follows, as a matter of pure arithmetic and not coincidence, that the marginal curve must cut through the average curve at the average's very peak. That single crossing point is no accident — it is the same logic that will later place the bottom of a firm's average-cost curve exactly where marginal cost passes through it.
Why this quietly builds your cost curves
Here is the payoff, and the reason this guide sits where it does in the rung. Flip the picture over. When a cook's marginal product is *high*, that cook produces a lot of extra pizzas for one paycheck — so the extra cost of each extra pizza is *low*. When marginal product *falls*, the same paycheck buys fewer extra pizzas — so the extra cost of each one *rises*. Marginal product and marginal cost are mirror images: as one falls, the other must climb. Diminishing returns in the kitchen *is* rising cost at the till.
Put one number on it. Say each cook costs $20 an hour. The second cook added 14 pizzas, so those came at about $20 ÷ 14 ≈ $1.43 of labour each. The sixth cook added only 3 pizzas, so those cost about $20 ÷ 3 ≈ $6.67 each in labour. Same wage, very different marginal cost — and the difference is nothing but diminishing returns wearing a dollar sign. This is the bridge to the next guides: the upward-sloping cost curves you are about to meet are not arbitrary. They are diminishing returns, told from the money side.
What the law does and does not claim
Be honest about the fences around this idea, because it is easy to overstate. First, the early *rising* stretch is real: the first few extra workers can specialise and fit the fixed plant better, so marginal product often climbs before it falls. The law is only about what happens *eventually*. Second, it is a short-run law by definition — it needs a fixed input. In the long run a firm can build a second oven, a bigger kitchen, a whole new branch, and then a different question takes over: what happens when you scale *every* input at once? That is returns to scale, a separate idea you must not confuse with this one.
Third, the law assumes the technology and the *quality* of the inputs stay put — every cook equally skilled, the recipe unchanged. That is the familiar all-else-equal clause, ceteris paribus, and the real world honours it only roughly. Bring in a clever new layout or a better oven, and that is no longer the same production function; you have changed the picture rather than disproved the law. This distinction matters at the level of a whole economy too: diminishing returns to capital alone is one reason lasting economic growth leans so heavily on new ideas and better technology, not just on piling up more of the same machines.