Up to this point, we only have determined that we want our economy to direct resources to their highest valued use for the purpose of maximizing society's happiness. But how can we determine the "highest valued use" of a resource? Consider the problem facing an economic planner in Cuba or the former Soviet Union. If a Soviet planner was trying to decide exactly where to build a new apartment complex, he might take several things into consideration. For example, he might do a study of housing conditions in various sections of Moscow. Based on that study, he might come to the opinion that new housing should be located in the most crowded part of the city or perhaps close to where a new factory is going to be built.

However the dilemma is not resolved when the location is decided. Next come questions such as: What kind of housing? Should the planner build family units or singles? Luxury apartments or just the basics? How big should the unit be? As you can see, this is a tremendously difficult decision--and an important one. From our previous discussion we decided that society is poorer if resources are not directed to their most valued use. If the planner directs bulldozers, architects, concrete and other construction resources to a part of Moscow where new housing is valued less than in another part of the city, he has done more than just made a mistake. He has hurt society. There is less happiness than what was possible. With all these difficult questions to answer, it is clear that the planner faces an overwhelming task.

Imagine how much more difficult the problem becomes when all of the resources of a country are considered. How many cars should the nation build? How many office buildings should be constructed? Should the farmers produce more vegetables, or the ranchers more beef? For the economic planner, there is simply no objective way to determine how best to use the resources that are available.

Does an objective way to measure "highest valued use" even exist? Let us consider this question in the context of an example. Let us suppose that a very wealthy, very eccentric old man meets his demise. In his will the old man directs the executor of the estate to divide his possessions among his relatives in the way that generates the most happiness. However, a codicil in his will dictates that the estate cannot be sold. Additionally, his relatives may not sell their inheritance. Whoever receives a particular allocation from this inheritance is restricted from trading or otherwise exchanging that allocation for other goods. What approach could the executor of the estate use to ensure that the maximum happiness was generated?

Consider first the approach of the planner. The executor could attempt to learn the relatives' preferences by undertaking a study. Based upon the results of his study, he would make a judgement about which relative would receive the greatest happiness from any given item. For example, in deciding who would benefit the most from having the old man's Mercedes, the executor would investigate which relatives have cars, how many cars they have, how old their cars are, etc. Suppose the results of his study showed that Uncle Morton already has three, very expensive luxury automobiles, while Cousin Ralph has no car and rides a bicycle to work every day. Then the executor could safely conclude that Cousin Ralph would get more happiness from the Mercedes than Uncle Morton. Or could he?

Maybe Cousin Ralph rides a bike to work each day because he enjoys the exercise. Or maybe he is a card-carrying member of Save the Planet!, and believes automobiles spoil the environment. In any case, if he received the Mercedes he would rarely use it. On the other hand, maybe Uncle Morton already has three luxury automobiles because he derives intense pleasure from luxury automobiles. Perhaps he has seven teenage sons who are always borrowing his cars. Either way, a fourth automobile would give him great joy. With lots of relatives and myriad goods, the informational requirements of his study would soon become overwhelming.

Consider another approach. Perhaps the executor says to himself, "These people know what they want. Rather than having me try to figure out their preferences, I'll just let them decide among themselves how to allocate the goods." Here he runs into another problem. How can the relatives compare their respective desires? Suppose the widow of the deceased says to the daughter-in-law, "You say you value the Mercedes a little, but I value it a lot. Therefore, I think I should get the car." What does "value a little" mean? What does "value a lot" mean? So the executor goes back to the drawing board, knowing that he must figure out some way of ranking preferences across relatives.

Then he gets a brainstorm. He decides to develop a questionnaire. Each relative would have to answer questions like: "On a scale from 1 to 10, how much do you value the Mercedes Benz?" "On a scale from 1 to 10, how much do you value the Louis XIV dining room table set?" The relative giving the highest rating would receive the item. The problem here is that it probably wouldn't take long for the relatives to realize that it pays to exaggerate on a questionnaire like this. There is no cost to lying about one's valuation. Even if a relative only believed a given item was worth a "7," he would always be better writing down a "10."

Consider a more sophisticated version of this questionnaire. Suppose the executor asks each relative to assign a rank to each item. Given a thousand items in the estate, each relative would rank order every item from their most preferred to their least preferred. While this represents an improvement on the previous approach, it also some obvious flaws. For example, what should the executor do when there are ties? Suppose an aunt, two nephews and a niece all put down the 3-D, virtual reality, home computer gaming system as their top choice? How does the executor decide which one gets it? And what does he do for the other three relatives? Does he give them their second choice instead? Suppose their second choice is somebody else's first choice?

Can you think of a better approach? It is our experience that, by now, most people come up with an alternative procedure that (i) utilizes all the information that is relevant for deciding the value of each item, (ii) allows the executor to compare preferences across individuals, and (iii) encourages truthful revelation by the relatives. And that procedure is an auction. The executor would allot each relative an equal number of tickets or coupons. Each relative would be free to bid as much or as little as they wanted, with items going to the highest bidder. In this way, each of the dearly departed's possessions would be allocated across the respective relatives. Note the attractiveness of this approach. The heirs are encouraged to think carefully about how much they value the goods, valuations can be compared across individuals, and no relative has an incentive to exaggerate.

Suppose Heir Number One outbids everyone else by bidding ten coupons for the Mercedes, with Heir Number Two being the next highest bidder at nine. The fact that Heir Number One has to give up ten of his coupons to obtain the car means that he has to think carefully about his valuation of the car. He has no incentive to bid more than it's really worth to him, because winning the bid means he loses the opportunity to purchase other goods. Further, because Heir Number One is willing to sacrifice more of other goods than Heir Number Two, we can say with some confidence that Heir Number One anticipates getting more happiness from the car. In other words, if the car goes to Heir Number One, we can be fairly confident that the most happiness possible will be generated.

We call this method of allocation the "willingness to pay" approach. Let's replace coupons now with dollars, rubles, pesos, etc., and let's CALL THE MAXIMUM AMOUNT OF MONEY A PERSON WOULD BE WILLING TO BID FOR A GOOD THEIR WILLINGNESS TO PAY FOR THAT GOOD. By allocating goods to those who are willing to pay the most, we allow individuals to reveal just how badly they want something. People who would receive only a little happiness from consuming a particular good would not be willing to pay very much for that good. In contrast, those who would receive a lot of happiness also would be willing to pay a lot. Thus willingness to pay provides one way to measure happiness.

The Soviet planner's dilemma is solved! To decide where resources would provide the most happiness, he just needs to discover how much money people would be willing to pay to enjoy those resources. If people in one section of Moscow are willing to pay more for housing than people in another section of Moscow, the planner's decision is now an easy one. But is willingness to pay really the best way to allocate resources? In the next chapter we consider this approach in greater detail.