Supply increases with price because more producers become interested in supplying as profits on the commodity increase. Demand decreases with price because fewer consumers can afford to purchase the commodity or are willing to substitute something else that costs less. When demand changes like this with price, demand is said to be elastic. However, if the commodity is a necessity and there are no acceptable substitutes for it, then demand will be inelastic -- it will remain about the same no matter what the cost.
Sometimes the acceptable substitute is the same commodity, but purchased from outside the normal economy. For example, if the commodity becomes too expensive to purchase in the normal market (because it is heavily taxed, for example), consumers may be able to obtain it in the black market, where the taxes are avoided. When there are no sources of the commodity outside the local economy, the local economy is said to be a closed economy. Where purchasers can go outside the local economy is said to be an open economy.
In the typical operant conditioning study, subjects (e.g., rats) are first deprived of food and then given opportunities to earn food in daily operant sessions. If they do not earn enough food in the operant chamber to maintain their weights, subjects are given supplemental food in the home cage.
If subjects are placed on fixed ratio schedules, then in effect they are being asked to "pay" for their food pellets with lever-presses. Raising the ratio requirement is equivalent to raising the "price" of the food. Economic theory would predict that demand for the food should decline as the price increases (the typical demand curve when demand is elastic). To some extent we do see this -- when the ratio becomes relatively large (e.g., FR-200), response rates drop and a point is soon reached when the subjects' demand for the food reaches zero -- they stop responding on the lever. This effect is even more apparent if subjects are given a second lever, where they can earn the same food at less cost. In that situation demand on the first lever will reach zero very quickly as subjects switch to responding on the other lever where the food is "cheaper." If the food offered for responding on the second lever is not the same as the food offered for responding on the first, and subjects prefer the latter, then subjects will be willing to continue responding on the first lever (for the preferred food) even if the price is somewhat higher, but even then they will switch to the second lever and the less-preferred food if the latter is sufficiently less expensive.
But why to subjects show ratio strain, and then stop responding, when there is only one lever and failure to respond means no food being earned? Perhaps it is because "free" food will be available after the session is over! According to this explanation, the food given in the home cage constitutes another source of food outside of the operant-chamber "economy." In other words, the way a typical experiment is run, the operant chamber represents an open economy, and we can therefore expect demand to be elastic. Ratio strain may be a phenomenon that appears only in an open economy. George Collier of Rutgers tested this idea by creating a novel situation in which a lab rat literally lived in the "operant chamber" 24-hours per day. The only way that the rat could obtain food was by pressing a lever, which delivered food on a ratio schedule. Under these conditions the ratio could be increased to as high as 1,500 responses per reinforcement, and the rat would keep banging away at the lever. At this ratio it took so long to complete the response requirement that subjects were spending virtually all of their time not spent sleeping, banging away at the lever. Ratio strain did not occur.
Collier's situation provided a closed economy with respect to food; there was simply no where else to turn to get it, and no acceptable substitute. The rats had to respond or starve. Under these conditions, economic theory predicts that demand should be inelastic -- the rats should be willing to pay just about any price in order to obtain the food. And that is precisely what Collier found.