The Competitive Firm: How a Price-Taker Decides

In T6 we built the cost side of the firm. You know how a producer’s costs behave as it runs faster or slower: marginal cost rises as output climbs, and average cost traces a U. What we never did was put a price next to those costs. A cost curve by itself cannot tell a firm how much to make, whether to make anything at all, or whether it will survive. For that we need the other blade of the scissors, the market price. Part A adds it.

We will study the simplest and in many ways the most important case: a seller so small relative to its market that it has no power over price. It must take the going price as a given and decide only how much to produce. Economists call such a seller a price-taker, and the markets full of them are what we usually mean by “competition.” This is the first of two installments on competitive markets. Part A stays inside a single firm and the industry it belongs to: how a price-taking firm picks its output, how thousands of them together generate a market supply curve, why competition produces the industry’s output as cheaply as possible, and why profits in such a market get driven toward an ordinary return. Part B puts this machinery to work on taxes, timing, and the constant churn of firms rising and falling.

A Price-Taker Faces a Horizontal Demand Curve at the Market Price

Start with a wheat farmer. Suppose wheat sells for $2 a bushel and the whole market trades ten million bushels, of which our farmer grows a thousand. Can he nudge the price up by withholding his crop? No. His thousand bushels are a rounding error against ten million; holding them back changes the going price by nothing anyone could measure. Can he charge $2.01? Also no. His wheat is identical to everyone else’s, so the instant he asks a penny more, every buyer walks to a rival and buys the same grain at $2. And he has no reason to sell for less, since he can already sell every bushel he has at the full $2.

So the farmer can sell as much or as little as he likes, but only at the one price the market hands him. Draw the demand curve facing him alone and it is not the familiar downward slope; it is a flat horizontal line at $2. That flat line is the signature of a price-taker: he takes the market price as fixed and chooses only a quantity along it.

What makes a seller a price-taker is not the raw number of competitors at any moment. It is two conditions working together. First, the product is homogeneous: buyers treat one seller’s units as interchangeable with another’s, so no one can command a premium. Second, entry is quick and cheap, so if a price ever rose far above cost, newcomers would pour in and compete it back down. A market can have only a handful of sellers today and still discipline each like a price-taker, as long as outsiders can jump in fast. Homogeneity plus easy entry, not a head count, is what strips a seller of power over price.

Set aside, for now, a different kind of seller: one that faces a downward-sloping demand for its own product, so that to sell more it must cut its price. Such a seller does not take a price; it searches for one. We study those price-searchers in a later topic.

For a Price-Taker, Marginal Revenue Is Just the Market Price

Because our farmer can sell any quantity at the same $2, the revenue from one more bushel is easy to find. Sell one more and revenue rises by exactly $2, with no hidden cost in the form of a lower price on everything else, because selling more does not move the price. The extra revenue from one more unit is what we call marginal revenue. For a price-taker, marginal revenue equals the price.

That is precisely what fails for a price-searcher, who must cut its price to move extra units and so earns less than the price on each new sale, since the cut bites into the units it was already selling. For that seller marginal revenue falls below price. Competition behaves so cleanly because, for a price-taker, marginal revenue and price are the same number.

The Firm’s Supply Curve Is Its Rising Marginal Cost

Now we can answer the first decision: at a given price, how much should the firm make? Run the marginal logic from T6. As long as one more unit brings in more than it costs, make it. The price-taker’s marginal revenue is the price, so the rule becomes: keep producing until marginal cost rises to meet the price. Stop where marginal cost equals price. Produce less and you are leaving units on the table that would have earned more than they cost; produce more and each extra unit costs more than it fetches.

Let me put numbers on it with a single producer’s cost schedule. These are the same cost figures we worked with in T6; here I have added what the firm does once a market price appears.

Daily rate of production	Marginal cost	Average variable cost	Average total cost
0	—	—	—
1	9	9.0	19.0
2	8	8.5	13.5
3	6	7.7	11.0
4	5	7.0	9.5
5	7	7.0	9.0
6	13	8.0	9.7
7	22	10.0	11.4
8	30	12.5	13.8
9	40	15.5	16.7
10	51	19.1	20.1
11	65	23.3	24.2
12	80	28.0	28.8

Suppose the market price is $25. Where does marginal cost meet it? At seven units. The seventh unit costs $22 to make and sells for $25, so it is worth producing. The eighth would cost $30 and still fetch only $25, so making it would shrink profit by $5. Seven units is the profit-maximizing rate, and notice we found it without ever consulting average cost. The marginal comparison alone pins down the rate.

Read the table the other way and you see the firm’s supply curve fall right out of it. At a price of $25 the firm offers seven units; at a higher price it would push to eight, nine, ten, climbing its marginal-cost schedule. The amount the firm is willing to supply at each price is read straight off its rising marginal cost. That is the deep result: above a certain floor, the firm’s marginal-cost curve is its supply curve. We will pin down that floor in a moment.

Shut Down or Keep Going Depends on Whether Price Covers the Right Cost

Finding the best rate is only half the decision. The other half is whether to produce at all, and a famous slogan does the work: marginal cost guides how much to produce, while average cost tells you whether to produce. Two different costs answer two different questions, and confusing them is one of the most common mistakes in economics. Marginal cost versus price sets the rate, as we just saw; to decide whether running at all earns or loses money, compare the price with average cost. But there are two average costs, because there are two time horizons, each with its own threshold.

In the short run, the firm has already sunk money into its building and machinery. That money is gone whether the firm produces or not, so it should not enter a decision about today’s output. What matters in the short run is only the cost that changes with production, the average variable cost. As long as the price covers average variable cost, every unit pays its own running expenses and chips in something toward the sunk investment, so the firm keeps going. In our table average variable cost bottoms out at $7, so at any price above $7 the firm keeps producing with the equipment it has; below $7 it cannot cover the cost of running and shuts down.

The long run is a different decision program, not the same one stretched out. When the equipment finally wears out and the firm must decide whether to replace it, the old investment is no longer sunk; buying new machinery is a fresh, avoidable cost. Now the relevant figure is average total cost, which includes the cost of capacity. It bottoms out at $9 in our table, so the firm reinvests and stays in the business only if the expected price covers at least $9. Below $9 it runs the equipment it has until it dies, then exits.

There is the floor we promised. The firm’s marginal-cost curve serves as its supply curve only above the lowest average variable cost, $7, since below that it produces nothing. And the lowest average total cost, $9, is the price below which it will not stay in the industry permanently. Same firm, two break-even prices, because the short run and the long run are two different production plans, each carrying its own cost.

Add Up Every Firm’s Supply and You Get the Market Supply Curve

One firm’s supply curve is its marginal cost above the shutdown point. A market is full of such firms. To get the supply curve for the whole market, ask at each possible price how much every firm would produce and add those quantities together. That horizontal sum, across all firms, is the market supply curve.

It slopes upward for two reasons. Each firm’s marginal cost rises with output, and a higher price also draws in firms that were sitting on the sidelines. Picture a price climbing from low to high: at first only the lowest-cost producers find it worth operating; as the price rises, each expands along its marginal-cost curve while new firms with higher costs clear their own break-even and switch on, adding fresh marginal-cost curves to the sum. Entry is the market’s own margin. The individual firm adjusts its rate; the industry adjusts the number of firms. Both push the same way, so the market offers more as the price rises.

Competition Pushes Every Firm’s Marginal Cost Toward the Same Number

Here is a quiet marvel that nobody plans. Every price-taker, whatever its circumstances, expands until its own marginal cost equals the one market price. They all face the same price, so they all end up with the same marginal cost, and that uniformity is exactly what it takes to produce the industry’s total output as cheaply as possible.

To see why, suppose it were not so. Imagine one more unit of some good costs $5 worth of resources in one firm but $6 in another. Then society is wasting something: shift resources from the costlier producer to the cheaper one and you get the same output for less. As long as marginal costs differ across producers, that cheap rearrangement is available. Resources flow out of the lower-valued use and into the higher-valued one until the marginal cost of one more unit is the same everywhere, and only then is no further saving possible. Competition performs that reshuffling automatically, with no planner computing anything, simply because every firm is independently chasing the same price. The total cost of the industry’s output is minimized as a byproduct of self-interested price-taking. It is the cleanest version of the old idea that competition serves the consumer without anyone intending to.

Survival, Not Conscious Optimizing, Is What Competition Selects For

I have been talking as though each owner equates marginal cost to price on a spreadsheet. Real managers rarely know their costs that precisely; they have hunches about whether to push output higher or pull it back, but the exact marginal cost of the next unit is usually a fog. So does the analysis collapse?

It does not, and the reason is one of the most important ideas in the course. The market does not require anyone to calculate the optimum; it only rewards those who land near it and punishes those who stray far. A firm that consistently produces too much, or carries costs that are too high, fails to cover them and is competitively winnowed out. A firm that happens to operate near the profit-maximizing rate survives and grows. Over time the survivors are the ones behaving roughly as if they had solved the problem, whether or not they ever consciously did. Profit-maximizing is better understood as a result selection produces than as a goal every manager pursues. The marginal conditions describe the firms that last, not the arithmetic each owner performs.

This reframes a familiar slogan. People say capitalism means consumer sovereignty, that consumer preferences alone decide what gets made. There is truth in it, since a firm that ignores what buyers want does not survive. But it is not the whole picture, because people choose as producers too, deciding what work they will and will not do and on what terms. What steers production is the interplay of those choices on both sides. “Individual sovereignty” is the more accurate phrase: consumers point the way, but they do not command production single-handed.

That winnowing has an unsettling public face. Certain industries, among them small groceries, bars, restaurants, coal mining, retail gasoline, textiles, and farming, are often dismissed as “sick” because most of the firms in them earn little or lose money while new firms keep pouring in about as fast as old ones fold. Critics read the churn as a defective, overcrowded market. It is nothing of the kind. Persistent losses need not signal a sick industry. An industry adjusting to falling demand shows broad losses across its firms with nothing pathological in the adjustment itself. Some owners knowingly accept below-market money returns because the work pays them in other coin, the satisfaction of growing orchids, acting, writing novels, or owning a ball club, so that a “loss” measured against what they could earn elsewhere is really a consumption choice. And in the winner-take-most fields, such as acting, writing, and professional sports, a large crowd of also-rans is exactly what every entrant is betting against.

The numbers confirm the pattern without indicting the market. Federal establishment data show that only about half of new private-sector businesses survive five years and only about a third survive ten, while roughly one in five closes within the first year, yet new firms keep arriving about as fast as old ones leave: in 2023 some 1.3 million establishments opened and about 1.2 million closed. Even the trades with the grimmest reputations are less deadly than legend claims. The often-repeated story that nine in ten restaurants fail in their first year is false; the best primary-data estimate puts first-year restaurant failure closer to one in six. An industry can carry chronic losses and a punishing exit rate and still be perfectly healthy, because the losses are the price of a bet that new entrants keep choosing to make.

Profits Turn Into Costs as Rivals Bid for Whatever Earns Them

Now to the engine that drives competitive profits toward an ordinary return, the genuinely distinctive part of this analysis. The usual story says competition “competes profits away” but rarely explains the mechanism, which is that profits get converted into costs.

Follow a single machine. An investor pays $10,000 for it, expecting 5,000 units of output at $1 per unit in other inputs, and hopes to sell the product for $3, which would just cover the machine. Then luck breaks his way and the product sells for $4. Each unit now clears $3 over the other-input cost, so over its life the machine throws off $15,000 of net earnings instead of the $10,000 it cost, and the owner is sitting on a $5,000 profit. Watch what happens to that profit. The moment everyone can see the machine earns $15,000, that is what it is worth; anyone who wants the business must pay $15,000, because that is what it will earn. The profit has been capitalized into the machine’s price, and whoever buys it at $15,000 now earns only an ordinary return, no windfall. The same happens if the resource responsible is a patent, a prime location, or a talented employee: rivals bid up its price or salary until the extra it earns is fully reflected in what it costs to obtain. The owner of that resource grows richer; the firm using it is left with a normal return.

The table tracks the machine as the product price moves.

Case	Product price	Net unit earnings (after $1 other inputs)	Machine’s value in use (= net unit × 5,000)	Cost of using the machine, per unit	Outcome
A	$4.00	$3.00	$15,000	$3.00	Invest; profit of $5,000 over the $10,000 paid
B	$3.50	$2.50	$12,500	$2.50	Produce; profit of $2,500
C	$2.00	$1.00	$5,000	$1.00	Keep producing; profit of $0
D	below $1	near zero	about zero	—	Terminate; the machine is worth scrapping

Two lessons live in that table. First, the cost of using the machine is not the $2 per unit you would compute from its $10,000 purchase price; it is whatever the machine is currently worth, spread over its output. When the product sells for $4, the true cost of using the machine is $3 a unit, because that is what you sacrifice by using it rather than selling it. The historical price you paid is irrelevant; only current value counts. That kills a stubborn fallacy: that new firms can undercut old ones because the old are “burdened” with obsolete, expensive equipment. They are not burdened at all. Old equipment is continuously revalued to whatever level lets it keep competing; if it can no longer earn its old keep, its value is rewritten downward, and at that lower value it competes on equal current footing with anything new. What a firm once paid for its capital is dead history.

Be precise about what these profits are. A genuine profit of this kind is an increase in wealth: someone created value worth more than the resources it consumed and for a while captured the difference. That is entirely different from the gain a firm collects by getting the government to block its rivals. Money squeezed out of consumers by restricting competition is not a profit in our sense; it is a transfer engineered by force, and competition does not erode it because competition has been forbidden. When we say competition drives profits to normal, we mean the wealth-creating kind, the kind that draws in imitators precisely because it is real.

One thread runs forward from here. When rivals bid up the price of a profit-earning asset, what they pay for is its stream of future earnings squeezed into a single price today. Collapsing a future flow into a present price is capitalization, and we will make its arithmetic explicit when we reach present value and capital later. For now the verbal version is enough: tomorrow’s profit gets baked into today’s asset price, leaving the new owner only a normal return.

Is Producing Where Price Equals Marginal Cost Socially Desirable?

Competition pushes each firm to produce where price equals marginal cost, and economists have long argued this outcome is, in a specific sense, a good one. We handle the argument in words, with no measured triangles.

At the competitive output, the last unit produced is worth at least its price to the buyer who takes it, since he was willing to pay that price. The price equals marginal cost, which is the worth of the other things society gave up to make that unit. So the last unit is worth as much to the person who gets it as the resources sacrificed to produce it. Push output lower and you skip units buyers value more than they cost; push it higher and you make units that cost more than they are worth to anyone. Producing right where price meets marginal cost is the point where the total personal worth created, across everyone, is as large as it can be. That is the sense in which the competitive result is efficient.

Be honest about what the argument assumes. It treats willingness to pay as the measure of a unit’s “worth,” and willingness to pay depends on a person’s wealth, so the efficiency claim accepts the existing distribution of wealth as its yardstick. That is not a hidden flaw, but it is a value judgment riding inside an apparently technical result, and you should see it. One footnote worth knowing: this argument was first worked out in detail by economists sympathetic to socialism, who hoped a planner could mimic it. It became one of the strongest cases for letting competitive markets and private property do the work instead.