A market where the goods are people
Until now you have watched supply and demand decide the price of coffee, of haircuts, of seats on a plane. This rung does something bolder: it points the very same machinery at *people*. The price we are after is the wage, and the venue is the labour market — a factor market, because labour is one of the factors of production a firm hires to make its output. The twist is that the two sides have swapped their everyday roles. Here the *firm* is the buyer, and *you* — the worker — are the seller, offering hours of your time.
That role-swap matters more than it first looks. When you buy a coffee, you want the coffee for itself — to drink it. But when a bakery hires a baker, it does not want the baker *for herself*. It wants the loaves she will bake and sell. Her labour is desired only as a means to something else. This is the single idea the whole rung turns on, so we give it a name in the next section — and once you have it, a wage stops being a number handed down from on high and becomes something you can almost compute.
Derived demand: pulled out of something else
Because no firm wants workers for their own sake, the demand for labour is a derived demand — it is *derived from*, pulled out of, the demand for whatever the workers produce. Nobody hires a welder out of fondness for welding; they hire welders because people want cars, pipelines, ships. So if you want to know what will happen to welders' wages, the first place to look is not the welders at all, but the market for the things welding makes.
This is derived demand, and it quietly explains a great deal. When the price of houses booms, the wages of carpenters and crane operators rise — not because building suddenly became noble, but because the value of what they build went up. When demand for printed newspapers collapsed, so did the demand for the people who printed them, however skilled. The demand for *people* rides on the back of the demand for *products*. Keep that image: the labour-demand curve is a shadow cast by the market for the final good.
What one more worker is worth: marginal revenue product
So how far does the firm push hiring? It applies the same margin-by-margin habit you used for output. For each extra worker it asks one question: how much extra revenue does this person bring in? That number is the marginal revenue product of labour — the worker's marginal revenue product, or MRP. It is built from two pieces you already know. First, how much *extra output* the worker adds (her marginal product). Second, how much *money* each unit of that output fetches when sold (the product's marginal revenue). Multiply them, and you get the extra revenue from hiring her.
MRP of labour = marginal product x marginal revenue
Bakery, selling each loaf at $4 (a price taker):
Bakers Loaves/hr Extra loaves MRP = extra x $4
------ --------- ------------ ----------------
1 20 20 $80
2 38 18 $72
3 52 14 $56
4 62 10 $40
5 68 6 $24
Hiring rule: keep adding workers while MRP > wage.
If the wage is $40/hr, hire 4: the 4th adds $40 (worth it),
the 5th would add only $24 (less than the $40 wage).Notice why the MRP column slopes down: each extra baker adds fewer loaves than the one before, because they crowd the same ovens — the diminishing marginal returns you met when studying the firm. The firm's rule, then, is beautifully simple: hire one more worker as long as her MRP exceeds the wage, and stop at the worker whose MRP has just fallen to meet it. Run that rule across every possible wage and you trace out a whole schedule of how many workers the firm wants at each wage. That downward-sloping schedule *is* the firm's demand curve for labour. The demand for people is literally the value of what the marginal person produces.
Your side of the deal: work versus leisure
MRP gives us the demand for labour. The other blade of the scissors is labour supply — the hours people are willing to offer. And here the model treats you, charmingly, as someone allocating a fixed and precious budget: the hours in a day. Every hour you sell to an employer is an hour you cannot spend sleeping, raising children, playing, resting. Economists bundle all of that non-work time under one word, *leisure*, and the choice to work is a choice about how to split your day. This is the work-leisure trade-off.
Seen this way, the wage is the *price of leisure*. Take an hour off and you forgo what that hour would have earned — its opportunity cost. So when the wage rises, leisure becomes more expensive, which tempts you to take less of it and work more. That is the intuitive direction, and over much of the range the labour-supply curve does slope upward: pay people more per hour and, on the whole, they offer more hours.
But honesty demands a wrinkle, and it is a famous one. A higher wage also makes you *richer* for any given number of hours — and when people grow richer, many choose to buy *more* leisure, not less, because rest and time with family are things we want more of as we can afford them. These two forces pull in opposite directions, and for some people, at high enough wages, the second wins: a raise leads them to work *fewer* hours. That is the celebrated backward-bending labour-supply curve. The plain lesson: "pay more, get more hours" is usually true, but not a law — the response depends on the person and the wage, and economists do not assume it always holds.
Where the wage settles
Now bring the two blades together. Labour demand (the firms' MRP) slopes down; labour supply (the workers' hours) slopes up. They cross at a single point, and that crossing sets the wage — exactly the market equilibrium you already know, just with people on the shelves. In the cleanest, most competitive version of the story, the equilibrium wage equals the marginal worker's MRP: in plain words, *in this idealised market a worker tends to be paid the value of what the last worker produces.* That is a striking and somewhat hopeful claim — and it is worth treating with care.
Care, because that tidy result leans hard on assumptions you should not swallow whole. It assumes many small employers competing for the same workers, so no single firm can push wages down. Where that fails — a lone factory or hospital that is the only real employer in town — the buyer of labour has power, a monopsony, and can pay *below* MRP. (Some economists argue this is one honest channel by which a minimum wage could raise pay without killing jobs, by clawing back wages a monopsony employer had been suppressing — though the size of any such effect is one of the most fiercely contested questions in the field.) The clean model also ignores that pay differs because *jobs* differ and *people* differ, the threads we pick up in the guides ahead.