Everyone knows that a surfeit of boys allows a girl to be more choosy. Excess demand for her companionship raises the "price" she can charge. And everyone knows that, with unchanged demand, excess supplies force prices downward: all farmers want bumper crops, but if every farmer harvests a bumper crop, silo operators get away with offering lower prices per bushel. The interaction of supply and demand determines the quantities traded and the prices at which buyers and sellers trade. By and large, the supply-and-demand model taught in a first-year college economics class is a good approximation of real life. Indeed, that is why we teach it!

Everyone knows that a surplus of boys allows a girl to be more choosy. Excess demand for her companionship raises the "price" she can charge.

But we make a number of assumptions. We assume, for example, that there are many buyers and many sellers so that the market is competitive. Where there is but one seller in the face of many buyers, the seller can be choosy – just as the girl. Or where a girl can sufficiently differentiate herself from other girls to make her appear more desirable, even unique, her price will rise. We also assume that markets are free of intervention from third parties, such as government. Restricting the number of tons of imported steel obviously limits domestically available supplies (at least temporarily) and drives up steel prices for steel users (autos, appliances, silver ware). That's why producers like to lobby their governments for trade restrictions, even as it hurts customers who end up paying higher prices – and leaves them with less money in their pockets to spend on other goods and services. We further assume that markets are private. For example, there is no particular reason why drivers' licenses need to be issued by government offices. The issuing of a license can be done by competing private firms. What government needs is merely the registration information and that is easy enough to pass on, just as private tax preparation firms pass on income-tax declaration information to the IRS.

Drivers' licenses can be issued by private firms. What goverment needs is  the  information, which is easy enough to pass on, just as tax preparation firms pass on income-tax information to the IRS.

We also assume that there are no spill-over effects to people unrelated to a particular market transaction. If you and a seller voluntary agree to exchange money for fast food, so be it. But if the resulting obesity epidemic leads to higher health-insurance premiums for me, I am essentially subsidizing your habits without any benefit to me (such as health-care premium reductions). Just as auto insurance is graded by driving record and risk factors, health insurance could similarly be graded. (It's easy enough to mandate an annual physical.) There can also be positive side-effects: if a beekeeper's bees fertilize a nearby orchard, the orchard owner and fruit buyers ought to pay the beekeeper for benefits received (as indeed orchard owners now do).

In the supply-and-demand model, we further assume perfect information on both sides of the aisle. But real markets are often characterized by imperfect information. For instance, if my car mechanic tells me that my car needs a $400 brake job, my lack of automotive knowledge makes me vulnerable. Possession of superior information pays off in gaining pricing-power or in preventing pricing-abuse. We also assume that a functioning contract resolution mechanism exists. If you and I write a labor contract, you need to be reasonably sure that you collect work for your money, just as I need to be reasonably sure to collect money for work. Trust in and credibility of contract enforcement is crucial to the proper functioning of markets.

There are more assumptions we make when we start out teaching supply and demand. Every time we relax one of the assumptions and allow more reality into our model, we explain more of the actually existing real-world markets, as the examples have shown. Another very important, and often unrecognized, assumption that has only recently been more thoroughly explored lies in the recognition that apart of autarky – total self-reliance in production and consumption – there are only two modes of economic activity. One is production and voluntary trade. The other is production and involuntary confiscation, i.e., appropriation. Thus, society needs to devote resources to the prevention, or least limitation of, appropriation. Many of the nasty (un)civil wars and resulting human and economic catastrophes we have seen in the 1990s, especially but not only in Africa, can be explained by the rise of appropriative, instead of productive, behavior. Societies that cannot adequately police themselves are prone to collapse.