Let’s look at choices in more detail. Every choice has two components.
Businesses have different objectives. Firms maximize profits, or, perhaps, shareholder wealth. In order to maximize their profits, they must also have chosen cost-minimizing levels of inputs for whatever output choice they settle on, so their choice process for selecting what output to sell also incorporates a choice of what inputs to combine. Politicians choose policies to maximize their own utility, perhaps by picking those policies that conform to their ideology, or alternatively by choosing to increase their electability. The two may not be much different.
But what determines the set of options, the menu of choices the chooser gets to select from? The simple fact that a choice must be made tells us that the options are limited, or constrained. We therefore think of the set of options open as one delimited by constraints. This is why we represent a choice formally as a problem of optimizing some objective function, such as a utility function, subject to a constraint or set of constraints.
The need to choose gives economics its definition of cost. The cost of choosing a particular option or bundle of goods is the value of the next best option, the best choice that was not selected. This best foregone option is called the opportunity costopportunity cost: the value of the best option not selected. of the option chosen. Notice that this definition of cost means that events which happened in the past, and which cannot be changed, do not count as costs. This means that a business that has already committed irretrievably to some expenditures should not take those expenditures into account in making its choices. For example, a company that has made a large investment in some technology, an investment that it cannot get back, should ignor the size of this investment in determining whether to press on either with more investment or with commercialization of the fruits of its research.
Costs such as these are called sunk costs. They do not matter to the choice problem for the simple reason that choices cannot affect them. For example, if a electric power company has invested in a large facility generating facility, but finds that once built, the additional costs per Kwh of electricity are higher than the costs of buying power from more efficient plants in the open market, the plant in question should be shuttered, no matter whether it cost a million or a billion dollars. Economics is relentlessly forward looking, simply because looking back makes for bad choices.
Suppose that Columbus decides to build a light rail system along I-71 to Polaris in order to reduce congestion. The city buys the right-of-way, spends a great deal of money in preparing to lay tracks and to build the stations along the way, but then Polaris gives up its fight with Easton for retail dollars and closes. Suddenly, the traffic load falls on I-71, but by now, the rail system is well along, and many of the costs have been incurred. Should the city finish it? If it does not, will it lose the, say, $1 billion that it has already invested?
The answer to the first question is maybe, and to the second question is no. Consider the choices available to the city. The first is to pay the additional cost to the city of finishing the line, and to reap whatever benefits the line will yield. The second is to abandon the line. In either case, the city is out its $1 billion. Since the amount already spent is gone either way, the choice made does not affect that amount. It is unaffected by the choice, and therefore irrelevant to the choice. What would be the difference if the cost was $2 billion because the mayor had made off with an extra billion as part of the deal? That is, what if politics in Columbus resembled politics in Chicago? That would make no difference.
The demise of Polaris would mean that the benefits to (completing) the project would fall. But the costs of completing the project would also be lower because of the work already done. To see if it would be worthwhile to complete it, we would need to compare the additional costs necessary to finish the project (not the total costs, including sunk costs) with the new, lower benefits. It still might make sense to complete the project, but notice that the decision needs to be entirely forward looking. And all “looking forward” means is refusing to pay attention to things you cannot change.
The same thing happens in less dramatic fashion elsewhere in the law. Most torts arise from accidents. Compensation for a crumpled car or a lost limb does not eliminate the cost of the accident. Payment of damages by a party that fails to live up to its contractual obligations does not mean that the obligations have been performed. Forcing a tobacco company to pay for illness caused by smoking does not undo the illness. The law seems often to look back, rather than forward. Economics, on the other hand, says “don”t cry over spilled milk.”
In this course, we will relentlessly retain the forward-looking perspective. In the case of murder, this means that the imposition of a punishment makes sense only if it is expected to yield reductions in future (not-yet-committed) crimes sufficient of offset its cost. The reduction in crime, if any, will be achieved by a mixture of keeping a person identified as a likely future wrongdoer from committing further crimes, and by indicating the cost to others of committing similar crimes. Similarly, compensation demanded of those who negligently cause accidents is analyzed by economists according to its effect on future accidents. If I know I must pay a substantial amount if I inadvertently torch my neighbor’s house while burning paint off my fence, I am likely to be more careful while doing so.
If the penalties imposed by the legal system are to shape conduct in the future in desirable ways, the people whose actions we want to shape must know something of the penalties they may be subjecting themselves to, as well as the probability those penalties will be imposed. Knowledge is not enough, however. They must be susceptible to having their behavior shaped by the prospect of penalties. This requirement takes us to a fundamental economic law, the law of demand. But before turning to responsiveness, we consider an application of opportunity cost to the different standards of proof applied in civil and criminal trials.
Why are standards tougher for criminal than for civil cases? One answer is that criminal punishments impose higher burdens on society than do civil cases. An incarcerated defendant imposes burdens on society, both because of the opportunity cost of that defendant’s time and because of the direct costs of the prison system such as guards and security apparatus. (What about food and shelter?) In contrast, penalties in civil cases are transfers—real resources are not consumed by the punishments, except by the functioning of the legal system itself.
But if the higher cost of penalties explains the differing standards, we still have a problem posed by our definition of opportunity cost. Suppose that it is easy and cheap for a plantiff to get a day in court in the event that a dispute arises over, say, a contract or an accident. Suppose also that no evidence need be presented by the plaintiff or defendant and that the jury, with nothing to go on, picks the willing party by flipping a fair coin. What is wrong with such a system?
The problem is that penalties are imposed on defendants whether or not they engaged in actions we wish to deter. The expected penalty is not a cost of engaging in a dangerous action, except to the extent that taking the action increases the probability of ending up in court. We do not get much deterrence unless we force plaintiffs to put up a substantial, if not overwhelming, case.