Economics and Ethics

Preferences over cars are dramatically different across individuals. Some people are perfectly content to drive low-cost brands of cars. Others prefer to drive high-cost brands.

In general, the choice that a consumer makes depends on one’s willingness to pay. For some, a car is just a way to get from point A to point B. There is no sense in buying a high-cost brand when a low-cost brand will do just fine. For others, the features or the quality differences (real or perceived) of the high-cost brand are worth the extra cost.

However, when economists refer to one’s willingness to pay, they are implicitly including one’s ability to pay. In the context of the way economists use the term, a person cannot say that he or she is willing to pay $1 million for a new Mercedes if the person does not have $1 million to spend.

Thus, a second reason why someone might choose the low-cost brand over the high-cost brand is that they cannot afford it. This might be true in an absolute sense that the car costs more than they earn. Alternatively, it could be that the cost of the high-end brand is deemed too high in terms of the foregone consumption that buying the car would require.

Economics allows us to make predictions about how cars will be allocated. In general, people with lower incomes or lower levels of wealth are more likely to buy low-cost brands. Note that this does not mean that higher-income people always drive the high-cost brands and low-income people always drive the low-cost brands. Rather, there is a tendency for people with lower incomes to drive the low-cost brands. Of course, as I wrote about in a previous post, the fact that high-income earners are more likely to drive high-cost cars might mean that high-cost cars signal status and that might actually increase the willingness to pay for people with lower incomes who value status.

Now, suppose that we are interested in determining whether the allocation of cars is good or bad. Economics doesn’t really tell us what is good or bad, but rather gives us tools for explaining how resources are allocated. The closest that economists can come to discussing whether something is good or bad is using the idea of Pareto efficiency. If something is Pareto efficient, then it is not possible to make one party better off without making another party worse off. In other words, an outcome is Pareto efficient when all gains from trade are exhausted. There is no one left whose willingness to pay exceeds the marginal cost of producing another unit of the good.

Car producers are price-setting firms. Each brand of car is different from others in terms of looks, style, and features. While some might prefer a particular brand, all else equal, this doesn’t mean that people will always choose their preferred brand. At a high enough price, the consumer will switch to another brand. In other words, these producers face a downward-sloping demand for their product. All else equal, higher prices mean a lower quantity demanded.

Suppose that producers charge a single, uniform price to all buyers. If so, the marginal revenue that the firm will generate is less than the price it charges. For example, suppose that the firm is charging $10 for its product and it sells 1 unit per day. Suppose that the firm could sell an additional unit per day, but would have to lower its price to $9. A a price of $10, the firm is generating $10 of revenue per day. At a price of $9, the firm generates $18 per day. The marginal revenue per day is $8, which is less than the price of $9. Why? Well, the firm gained $9 in revenue from selling the second unit of the good, but lost $1 in revenue from selling the first unit at a lower price.

This is important for the following reason. If these producers are profit-maximizing, they will choose a price such that marginal revenue equals marginal cost. However, since the price is always greater than marginal revenue, this means that the marginal buyer is paying a price above marginal cost. Pareto efficiency requires that price equals marginal cost. If the price is above marginal cost, then there will likely be some buyer who is willing to pay a price above the marginal cost, but below the uniform price set by the firm. This isn’t Pareto efficient. There are possible gains from trade that are unexploited. Thus, one might say that this uniform pricing allocation is “bad.”

To solve this problem, producers need to figure out how to charge their existing customers the uniform price while simultaneously selling at a lower price to these other customers. To return to the car example, they can sell cars at different prices to different people in an attempt to sell cars at prices that are equal to the individual consumer’s willingness to pay. This seems to be what happens in the market for cars. Car dealers put stickers in the window that itemize the cost of the vehicle and list a price. However, many people spend a long time at car dealerships haggling over the price. Everyone ends up paying a different price and this works out better than the alternative of uniform pricing because fewer gains from trade are foregone.

The result of this process is that people pay different prices for the same car and also choose different cars based on the differences in prices. The extent to which economics can describe this as good or bad is limited to a discussion of Pareto efficiency. The fact that people pay different prices for the same car is likely “good” in the sense that it gets society closer to exhausting the gains from trade. The fact that people choose different cars based on the differences in prices is likely also “good” because consumers are going to try to get the most consumer surplus (what they are willing to pay less what they actually pay) that they can. Thus, if a particular consumer chooses a particular car, one can reasonably assume that this was better than the alternatives. Overall, whether the outcome in this market is good or bad depends on whether there are possible gains from trade that are unrealized.

Brian’s post about pollution was designed to get people thinking about Pareto efficiency. Renters can choose from a variety of different places to live. These places differ in size and location. Some people are willing to live close to pollution (and bear the cost of pollution) for a lower monetary cost in terms of rent. Brian wrestled with Armen Alchian’s point that a magic wand that eliminated pollution at zero cost might not be “good.” This seems wrong. If society could eliminate pollution at zero cost, isn’t this necessarily good? Alchian’s answer was no. The renters in areas that formerly had a lot of pollution would see an increase in their rent. They would face a higher monetary cost of renting, but a lower cost of pollution. Whether the person was better off would depend on whether the increase in rent was greater than the cost of pollution to the current renters. If so, then they are worse off. What Alchian was referencing here was the notion of Pareto efficiency. If one group is worse off, then there hasn’t been a Pareto improvement.

My fellow economists who are reading this might argue that Pareto efficiency isn’t the only criteria for addressing welfare. These economists are both correct and wrong. They are correct in that there are other ways to evaluate whether policies are good or bad. However, they are wrong in the sense that those other criteria necessarily involve value judgments. They take the economist away from economics and towards ethics.

For example, many economists would argue that if the benefits of a policy exceed the costs, then the policy is “good.” This is a utilitarian argument. That is not an economic argument. If a particular policy benefits me $500 and costs you $300, a utilitarian would say that this is “good.” Should an economist say this is good? Perhaps, if they are a utilitarian. However, this policy is a Pareto improvement (i.e., “good” in an economic sense) if and only if I compensate you $300. This leaves me with a $200 gain and you are no worse off than you were before the policy.

This might seem like a minor point, but it is critical to understand the boundaries of where economics ends and value judgments begin. This comes up all the time in discussions of international trade. Economists generally agree that the benefits of eliminating a tariff exceed the costs. However, the costs are disproportionately borne by those who might lose their job or experience permanently lower wages. Even if this is the case, the policy of eliminating the tariff can be “good” in a Paretian sense if the losers are compensated by the winners. Yet, many economists argue that there is no reason to compensate the losers. Perhaps there is not, but that is a utilitarian argument and not an economic argument. These people are making a value judgment.

This (supposedly) gets even murkier when we venture outside of markets. People seem to conclude that we cannot use the same logic when prices aren’t involved. However, even where prices are absent, costs remain; and costs are all that matter.

To understand what I mean, consider that, in my car example, no objective criteria were used in evaluating the equilibrium. I argued that an economist can only rely on criteria related to the gains from trade. To determine whether the allocation of cars was good or bad, I made no reference to whether the car had power steering or third-row seats, or a particular minimum gas mileage. Instead, it was assumed that consumers would sort themselves given the prices they observe, their budget constraint, and their preferences for the various features and characteristics of the cars. Given the consumers’ subjective valuations, one can examine whether this outcome is "good” or “bad” by comparing the potential gains from trade to the realized gains from trade.

Yet, when it comes to things like public schools, whether schools are deemed a success or a failure is entirely judged by objective criteria. What is the graduation rate? What is the average test score? Conceivably this has something to do with the fact that the services of public schools are not sold at a particular price in a market. But, from an economic perspective, how is the evaluation of schools any different from the car example or the pollution example?

If someone buys a Kia instead of a BMW, economics tells us that the cost of the BMW is too high to this person relative to the Kia to justify its purchase. That might be related to preferences or it could be related to their budget constraint. On average, economics would tell us that higher-income people will tend to buy BMWs and lower-income people will tend to buy Kias, but it is not strictly true that high-income people buy BMWs and low-income people buy Kias. Furthermore, to determine whether this allocation is “good”, economists don’t assign objective criteria to cars and then compare whether those who purchased a Kia are getting the same value as a BMW.

Furthermore, when people are willing to rent in high-pollution areas, we say that they are willing to pay the cost of higher pollution for the lower monetary cost of rent. Conversely, people who rent in low-pollution areas clearly prefer to pay a higher monetary cost to get a lower cost of pollution. Again, lower-income people will be more likely to live in areas with higher pollution and higher-income people will be more likely to live in areas with less pollution, but it is not strictly true that high-income people will live in low pollution areas and low-income people will live in high pollution areas.

When it comes to schools, economics tells us that bad schools are sort of like pollution. Some people will be willing to pay higher rent to live in an area with better schools. Other people are willing to bear the cost of a lower-quality school in exchange for cheaper rent. Furthermore, the choice of school district might also depend on the value of the parents’ time. If a parent has the time to spend with a child to supplement the teaching and help with homework and study for tests, the parent could choose the cost of foregone leisure (and even lower-quality schools) if it means lower rent.

As in any other market, it is likely that parents are a mixture of people with different objectives. Some parents are willing to move into the expensive neighborhood with high property taxes if it means access to better schools while sacrificing other expenditures. Others might live in the expensive neighborhood for other reasons and the better school is just a bonus. Some parents might choose to live in a neighborhood with “bad” schools and supplement the in-class learning with teaching at home to compensate for lower-quality schools. They might choose this option because they have a greater abundance of time than money and this allows them to spend money on other things that time cannot buy. Others might choose the “bad” schools simply because they don’t put much value on education.

All of this is complicated by a principal-agent problem since the student gets the primary benefit of the education, but parents bear the costs. I’ve written about this problem before.

Regardless, economics is fundamentally about subjective value. Some people certainly look at BMWs and think the people who buy them are wasting their money when a perfectly good Ford or Toyota will get the job done at a substantially lower cost. Perhaps some parents look at the “good” schools in the same way. Parents might have different definitions of quality and different valuations of various characteristics. Economics helps us to understand the decisions people make and how people sort, but the extent to which economics can tell us about whether such outcomes are good or bad is limited to whether there are unrealized gains from trade.

For evaluating what is good and what is bad, economics offers Pareto efficiency. We can analyze outcomes and try to determine if there are gains from trade to be had that are foregone for some reason. Beyond that, evaluating actual outcomes requires value judgments that economics alone cannot provide.

Of course, none of this is to say that there is no role for value judgments in economics. On the contrary, value judgments are necessary for any type of policy analysis. In fact, choosing Pareto efficiency as the evaluation tool is itself a value judgment. Nonetheless, my argument is that it is always important to distinguish between what economics has to say and what a particular economist has to say.