Module 12: Input Markets

 

THE POLICY QUESTION
SHOULD THE STATES OF OREGON AND NEW JERSEY PROHIBIT SELF-SERVICE GAS?

Only two states in the United States do not allow self-service gas stations: in Oregon and New Jersey, customers are not permitted to pump their own gas. Originally, this policy was driven by safety concerns—gasoline was considered a hazardous substance due to its flammability and the adverse health effects of touching it or inhaling fumes. However, advances in pump technology have made the self-pumping of gas safe enough that forty-eight states and the District of Columbia have allowed self-service gas for decades. Efforts to repeal the prohibition in Oregon and New Jersey have failed, and one of the key arguments from groups that defend the policy is that the repeal would lead to job losses. Economists often claim that policies that mandate employment are inefficient and cause a drag on the economy itself.

In this chapter, we will look at input markets and study labor markets in particular. Labor is an input in retail gas, so we will be able to study this policy in depth and think about the economic implications. We can also use what we have learned so far about product markets to think through the implications for consumers.

EXPLORING THE POLICY QUESTION

1. Do you think prohibiting self-service gas is a good policy? Why or why not?
2. Do you think that policies such as this one help ensure that there is adequate employment?

12.1 INPUT MARKETS

Learning Objective 12.1: Explain how changes in input markets affect firms’ cost of production.

In chapter 6, we learned that when firms produce a good or service, they do so by combining various inputs. These inputs, also known as factors of production, have a price. For example, we learned that capital has a price called the rental rate and labor has a price called the wage. But inputs can be almost anything, like aluminum for auto manufacturers, cocoa beans for chocolate makers, or fresh produce for a restaurant. When firms buy these inputs, they do so in goods markets and act as consumers just like those in our demand models. In that way, input markets are just the same as the markets we have already studied. In most cases, the markets work like the goods markets we have studied, but there are some important exceptions, and we will study two in this chapter.

The first exception comes from the input used by virtually every producer, labor. Labor markets have particular features that make them different than normal product markets. The first is that labor is not a good or service supplied by a firm but the physical and mental effort exerted by individuals in exchange for a wage. There are both physical constraints on the supply of labor—a person cannot supply more than twenty-four hours’ worth in a day—and constraints that come from individuals’ decisions to consume leisure time as well as tangible goods and services.

The second exception comes from the fact sometimes suppliers of inputs supply a single company. Thus the purchasing company acts as a monopsonist: a single buyer for goods or services sellers. Monopsony is the term used to describe a market in which there exists only one buyer for a good. Monopsonies are rare, but they do exist. Examples are the market for sophisticated weaponry for which there is only one legal buyer, the Department of Defense, and the market for labor in “company towns” where one firm employs most of the workers in the town.

Before studying the exceptions, let’s take a look at a typical input market that acts as a normal goods market, with many sellers and many buyers. Consider a company like Dell that sells PC computers fully assembled and ready to use. This company may make component parts themselves but also likely purchases component parts from other suppliers as well. Let’s suppose that Dell buys computer hard drives (HDD) from various suppliers. There are many manufacturers of computer hard drives and many potential buyers, including individuals who wish to build or upgrade their own computers, so the computer hard drive market is very typical of a standard goods market.

Note that the hard drive market is both an input market, as hard drives are inputs in the manufacture of Dell computers, and a final goods market, as hard drives are also final purchases of consumers. Some goods, like raw iron, are almost exclusively inputs because there is no demand for raw iron as a final consumer good. A good that is used as an input to produce other goods is called an intermediate good. Other goods, like a pair of denim jeans pants, are purchased by the end user. Such a good is called a final good. And some goods, like computer hard drives, are both intermediate and final goods.

Manufacturers like Dell purchase computer hard drives in product markets along with regular consumers, though they typically buy direct from the manufacturers and not through retailers, as do consumers. Regardless, the price they pay for their hard drives is set by the hard drive goods market, as illustrated in figure 12.1, which shows the market for one particular type of drive: the one-terabyte HDD (1TB HDD).

Figure 12.1 The market for one-terabyte computer hard drives (1TB HDD)

In figure 12.1, there are many suppliers of hard drives and many demanders. The demanders are both manufacturers and consumers. Since this is the market for wholesale 1TB HDDs, consumers are represented by retailers like Best Buy, Amazon, and New Egg. The price of a 1TB HDD is determined by the market as $P^{\ast }$. This $P^{\ast }$ is an input price for computer manufacturers, just like the rental rate on capital, $r$, and the wage rate for labor, $w$, as we studied in chapter 6. The same market dynamics of goods markets apply here: when demand or supply changes, prices will change as well. For example, if new manufacturers of 1TB HDD enter the market, the price for these goods will fall. If there is a new substitute for 1TB HDD, like solid-state hard drives, demand could fall, which would also lower the price of the good.

Price changes in the input markets will affect the cost conditions of firms, as we saw in chapter 6, and will, in turn, affect the supply of the final good. As input prices rise, the supply curve in the final good market shifts to the left, and the price of the final good rises. As input prices fall, the supply curve of the final good market shifts to the right, and the price of the final good falls.

As noted above, the typical behavior of input markets changes when the input is labor or when the buyer of the input is the only buyer. We study these two instances in the next two sections.

12.2 LABOR SUPPLY

Learning Objective 12.2: Describe how individuals make their labor supply decisions and how this can lead to a backward-bending labor supply curve.

Almost every final good produced has some labor input involved, if not as part of the manufacturing process then as an ancillary contribution through administration, marketing, sales, and so forth. For this reason alone, labor has an outsized role in production and deserves special attention. Another particular aspect of labor that requires special attention is the fact that it is an input provided by individuals making optimal decisions about work and leisure.

The labor market is where labor prices, or wages, are set and is defined by the same supply-and-demand dynamics as other input markets. The supply of labor is determined by the labor-leisure choice of individuals. Individuals think about the trade-off between working more and earning more money, which can be used to purchase goods and services, and working less and having more time to enjoy leisure activities and perform chores that are necessary but for which there is no monetary compensation.

The labor-leisure choice is the starting point for thinking about labor markets. We can model this choice just as we do consumers deciding between bundles of two different goods. Both leisure and the income from labor are goods that reflect standard preferences: individuals would prefer more of both, and a mix of both income and leisure is preferable to too much of one or the other. Thus we can describe an individual’s preferences by a utility function of income ($I$) and leisure ($l$):

$U=U(I,l)$

Individuals cannot spend more than twenty-four hours per day in either activity. We can write down a constraint that describes this as

$H=24-l$,

where $H$ is the hours spent working in a day and l is the amount of leisure consumed in a day. For a given hourly wage rate, $w$, the income earned by a consumer, $I$, is given as

$I=wH$.

Recall that in chapter 4, where we studied the consumer choice problem, we took income ($I$) as given to the consumer. In this discussion, we will see that $I$ is a result of a decision the individual makes about how much labor to supply to the labor market.

Figure 12.2 shows the optimal choice between leisure and income for an individual, Asha, and how it is determined. Asha’s income is measured on the vertical axis, and the amount of leisure hours per day she consumes is measured on the horizontal axis. Since there are only twenty-four hours in a day, Asha cannot consume more than twenty-four hours of leisure. Note that the less leisure she consumes, the more time she spends on labor and the more income she earns.

Figure 12.2 shows Asha’s budget constraint line in black. If the wage rate is $w_1$, the budget constraint has the slope of $-w_1$. This is because for every extra hour of leisure she consumes, she loses $w_1$ in income. The highest indifference curve Asha can achieve is given as $U_1$ and is shown in red. Maximizing her utility at the wage rate of $w_1$ yields a choice of $I_1$ in income and $l_1$ in leisure.

Figure 12.2 The optimal labor–leisure choice problem

Now, what happens when the wage rate increases to $w_2$? Asha’s new budget constraint, shown in blue, lies above the old budget constraint. This is because at a higher wage, Asha gets more income for every hour spent working or not consuming leisure. The highest indifference curve she can achieve at this budget constraint is given as $U_2$. How does Asha adjust her consumption? Since the opportunity cost of leisure has increased, she gives up more income when she consumes leisure and she consumes less leisure. Maximizing her utility at the wage rate of $w_2$ yields a choice of $I_2$ in income and $l_2$ in leisure.

From the solutions to this optimal labor–leisure choice problem, we can derive the demand curve for leisure for Asha (figure 12.3).

Figure 12.4 An individual’s labor supply curve

Asha’s labor supply curve, as depicted in figure 12.4, is an upward-sloping curve. The curve shows that the effect of an increase in the wage is an increase in hours of labor supplied from $H_1$ to $H_2$. This change, however, is the net effect of both the income and substitution effects, which we studied in chapter 5.6. The income effect refers to the fact that as the wage rises, Asha has more money to spend on both leisure and other goods, and if the goods are both normal, she will want to consume more of both. The substitution effect refers to the fact that as the wage rises, leisure becomes more expensive relative to the consumption of other goods, and Asha will naturally substitute away from leisure and toward the increased consumption of other goods, as enabled by more work at the higher wage. If the substitution effect dominates the income effect, then her labor supply will be upward sloping, as in figure 12.4.

Suppose, however, that the income effect dominates the substitution effect. In this case, Asha will choose to consume more leisure as her wage rises, and her labor supply curve will be backward bending. This might be more likely as wages rise high enough that individuals can satisfy much of their material consumption desires and increase their consumption of leisure. Put another way, there may be a point where increased income causes individuals to actually choose to work less so that they can have more time to enjoy the consumption of their goods, services, and leisure. Figure 12.5 shows three different wages and the corresponding leisure choice for each. Figure 12.6 illustrates the resulting labor supply curve, which starts positively sloped but then bends backward at higher wages.

Figure 12.5 The backward-bending labor supply curve: the labor-leisure choice

Figure 12.6 Labor supply curve slopes upward at lower wages and then bends backward at higher wages.

When the wage rises from $w_1$ to $w_2$, Asha decides to work more hours and consume less leisure. However, when wages rise from $w_2$ to $w_3$, Asha decides that her income is high enough that she would like to consume more leisure and work less. That is, her new wage allows her to consume more of both goods (leisure and all other goods). In this case, the income effect, which causes her to want to consume more of both goods, outweighs the substitution effect, which causes her to substitute away from the relatively more expensive good, leisure, and toward the relatively less expensive good, consumption of other goods.

This analysis shows that a backward-bending labor supply is theoretically possible, but do such labor supply curves actually exist? Only empirical research can answer this question. But there are some assumptions that are not quite accurate for many workers, the first being the characterization of the labor-leisure choice. Most workers cannot choose independently how many hours per day they work. Many jobs have specified work hours, and in the United States, many jobs are forty hours a week with very little room for adjustment. So when employers raise wages, there may be no way for employees to adjust their hours in response. Empirical research suggests that for men, this rigidity in hours might be very important, as their labor supply curves have been estimated to be nearly vertical.

12.3 COMPETITIVE LABOR MARKETS AND MONOPSONIST LABOR DEMAND

Learning Objective 12.3: Explain how monopsonist labor markets differ from competitive labor markets.

Another way that input markets can be different than a typical goods market is when there is only one buyer of an input. Monopsony is the situation where there are many sellers but only one buyer of a good. Consider an advanced weapon system that by law can only be sold to the government or a factory town where almost all the resident workers are employed by a single factory.

The case of professional American football players in the United States is a good example. The National Football League (NFL) is the only major professional American football league in the world, and while there are a few other options, like the Canadian Football League, they pale in comparison to the NFL. In labor market terms, the NFL acts a lot like a single entity with strict rules about hiring that prevent teams from competing with each other for new talent. Each year there are many sellers of labor, the football players, and one buyer, the NFL.

Competitive Firms’ Employment Behavior

How are monopsonists different than competitive firms in a labor market? To answer this, let’s first look at competitive firms’ hiring decisions. In a competitive labor market, each firm has no influence on the equilibrium market wage, so no matter how many workers or worker hours a firm employs, the wage will remain constant. Thus a firm’s decision is to continue to employ workers as long as the value of their marginal product of labor ($M{P}_L$) is greater than the wage. Recall that $M{P}_L$ is the extra output achieved from the addition of a single unit of labor. For our discussion here, the unit is an hour of work. The marginal revenue product of labor ($MR{P}_L$) is the value of the marginal product of labor—it is the extra revenue a firm receives for an additional unit of labor. Therefore, the $MR{P}_L$ is the product of the marginal product of labor and the marginal revenue, $MR$, that accrues to the firm from one additional unit of output:

$MR{P}_L=MR\times M{P}_L$

A firm will continue to add additional hours of labor as long as that additional labor adds positively to profits. The cost of an hour of labor is $w$, and the benefit is precisely the $MR{P}_L$. So as long as $MR{P}_L>w$, the firm is adding to its profits. If the $MR{P}_L<w$, the firm can increase its profits by reducing the number of labor hours it uses. The firm maximizes its profits from its employment decisions only at the point where the $MR{P}_L=w$. So we can characterize the firm’s employment rule as

$MR{P}_L=w$

In a competitive output market, the $MR$ is equal to the price that each unit sells for in the market, or the market price, $p$. So $MR{P}_L=p\times M{P}_L$. The firm’s employment rule becomes

(12.1)$p\times M{P}_L=w$

Dividing both sides by $M{P}_L$ yields

$p=w/M{P}_L$

We know from chapter 9 that profit maximizing firms produce to the point where marginal revenue equals marginal cost. Marginal revenue is $p$, as we just discussed, so the profit maximizing rule is

$p=MC$

Putting together the employment rule and the profit maximization rule, we get

$p=w/M{P}_L=MC$

Recall from section 6.3 that as the firm adds incremental units of labor, the $M{P}_L$ increases up to a certain point and then begins to fall due to inefficiency. Beyond this point, the firm must reduce units of labor in order to increase $M{P}_L$. We can combine this fact with the expression of labor demand in equation (12.1) to explain the firm’s employment behavior.

Specifically, with $p$ fixed from the final goods market, as $w$ rises, the firm will reduce worker hours in order to increase $M{P}_L$ and so maintain the equality in (12.1). We can see this in figure 12.7, which shows in black the labor demand curve, $D$, when price is $p$. At wage $w_1$, the firm will employ $L_1$ hours of labor; when wages rise to $w_2$, the firm will reduce employment to $L_2$. This is a shift along the labor demand curve as wages change.

Figure 12.7 The labor demand curve and wage and price changes

What happens when the price of the final good changes? Well suppose the price rises from $p$ to $p^{\ast }$. This means that the $MR{P}_L$ rises as well, and so temporarily, the $p\times M{P}_L>w$. The firm can now add more workers to lower the $M{P}_L$ until the equality of the $MR{P}_L$ to $w$ is restored. We see this in figure 12.7 as a shift in the labor demand curve from $D$ (in black) to $D^{\ast }$ (in blue). Now at wage $w_1$, the firm will employ $L_1^{\ast }$ hours of labor, and when wages rise to $w_2$, the firm will reduce employment to $L_2^{\ast }$.

To summarize the results of wage and price changes on the labor demand curve,

• a change in wage results in a shift along the labor demand curve, and
• a change in price results in a shift of the labor demand curve.

Wages and Labor Market Equilibrium

How is the wage amount set in a competitive labor market? To find labor market equilibrium, we must first derive the market supply of labor and the market demand for labor.

The market supply of labor is simply the sum of the labor supplied by all individual workers. Figure 12.4 showed an individual labor supply curve; the market labor supply curve is illustrated in figure 12.8. Labor hours, $L$, is simply all the individual work hours per day, $H$, for every potential worker. As the wage increases, the number of workers and the number of hours each worker is willing to work increase as well, and the result is an upward-sloping curve.

Figure 12.8 The market supply of labor

We find the market demand for labor by summing up the individual firms’ labor demands. Once we have the market demand for labor, we can graph it together with the market supply of labor, as shown in figure 12.9. Doing so reveals the labor market equilibrium, identical to the equilibrium in other goods markets, and the resulting equilibrium wage rate, $w^{\ast }$, and total employment, $L^{\ast }$.

Figure 12.9 Labor market equilibrium

As figure 12.9 shows, we call the market demand for labor $D_L$ and the market supply of labor $S_L$. Where they cross is the equilibrium point, and the wage and labor hours at that point are the equilibrium wage rate, $w^{\ast }$, and equilibrium employment, $L^{\ast }$, respectively.

Monopsony Employment Behavior

In a competitive labor market, the wage amount is set by the market equilibrium of labor supply and demand, and the firm must pay that wage in order to attract workers. But when the firm is a monopsonist in the labor market, it has the power to influence the wage amount, $w$, and does not take it as fixed.

A competitive firm takes the wage as given and can hire as many work hours as it likes at that wage. We call the marginal expenditure ($ME$) the extra cost of hiring one more unit of labor. The $ME$ for a competitive firm of an additional hour of labor is simply the market wage.

A monopsonist must also pay a wage to attract workers, but this wage depends on the labor supply curve. For example, suppose that at a wage of $20 per hour, a monopsonist firm attracts one thousand hours of workers per week for a total weekly wage bill of $20,000. Now suppose the price of its product has increased, and it wants to increase production. To do so, the firm will have to hire more work hours by raising the wage.

Suppose that at $25 per hour, the firm now attracts twelve hundred hours of workers. The firm’s total wage bill has now increased not just by the extra two hundred hours of work at $25, or $5,000, which would increase its wage bill to $25,000. Rather, it now pays for all of its labor hours at $25. This increases the wage bill to $30,000 ($1,200\times \$25$). So the marginal expenditure, $ME$, of an additional hour of labor is not just the additional wage the firm has to pay for that hour of work but the additional wage the firm has to pay for all hours of work.

Figure 12.10 illustrates the marginal expenditure curve for a monopsonist firm. The $ME$ curve lies above the market supply of labor curve because of the fact that the firm cannot pay for one additional hour of work at a slightly higher rate than the rest. It must pay the same wage for all hours of work. The firm is therefore setting its $MR{P}_L$ equal not to the wage but to the $ME$. We can state this insight about labor market equilibrium in the case of a monopsonist firm as

$MR{P}_L=ME$.

Figure 12.10 Monopsony and the marginal expenditure curve

For a monopsonist firm, the $MR{P}_L$ curve is the labor demand curve. The firm’s optimal employment decision is where $ME=D_L$. This occurs at $L^{\text{Monopsony}}$. To attract this amount of labor, the firm must pay $w^{\text{Monopsony}}$—the point on the supply curve that yields precisely $L^{\text{Monopsony}}$. We can see from figure 12.10 that the monopsonist will hire fewer hours of labor ($L^{\text{Monopsony}}$) than would occur in a competitive market ($L^{\ast }$) and pay a lower wage ($w^{\text{Monopsony}}$) than in a competitive market ($w^{\ast }$).

12.4 MINIMUM WAGES AND UNIONS

Learning Objective 12.4: Show the labor market and welfare effects of minimum wages in a comparative statics analysis.

In a monopsonist labor market, the buyer of labor—the firm—has some power to set wages rather than accepting a market-driven rate. In other contexts, the sellers of labor—workers—may enjoy an advantage in determining wage rates, employment levels, or both. These advantages can be achieved through collective bargaining, for example, or public policy approaches, such as setting a minimum wage.

In the United States, it is illegal to pay workers below the federally mandated minimum wage, currently set at $7.25 an hour. Many states have set minimum wages above this level; in 2015, Washington was the state with the highest minimum wage, at $9.47 an hour. There are many reasons for these policies, but the most common rationale, often expressed by those who champion even higher minimum wages, is that wages are too low and do not represent a “living wage”—a wage high enough to afford the basic necessities of food, shelter, and clothing. In other words, minimum wages are a policy tool to address poverty.

To analyze minimum wages, we start by recognizing that they are price floors in the labor market, similar to the price floors in goods markets that we studied in chapter 11. A minimum wage set above the market equilibrium wage moves the market leftward on the market demand for labor curve, as figure 12.11 shows. As a result, fewer employers are able to enter or remain in the market, and existing employers demand fewer hours of labor at the minimum wage than at the market equilibrium wage, $w^{\ast }$. At the same time, more workers are interested in entering the market, and existing workers are willing to work more hours at the higher minimum wage than at the market equilibrium wage. This difference between the number of hours that workers would like to work ($L^{\text{supplied}}$) and the number of labor hours that employers demand ($L^{\text{demanded}}$) leads to an excess supply of labor at the minimum wage, as shown in figure 12.11.

Figure 12.11 Minimum wage in a labor market

Economists describe the difference between the employment level at the equilibrium wage, $L^{\ast }$, and the employment level at the minimum wage, $L^{\text{demanded}}$, as the disemployment effect: the amount of employment lost due to the minimum wage. We refer to the difference in the amount workers would like to work at the minimum wage and the amount they actually do work as unemployment. Unemployment occurs when there are people who would like to work at a given wage but are unable to find employment.

Unions are labor groups that bargain collectively for wages for a group of workers in a company. Their objective is to leverage collective bargaining to achieve higher wages than would have occurred in the individual labor market. In this sense, then, we can think of the result of unions in a similar way to minimum wages set by governments: they create a price floor in labor markets.

How do wage floors affect the welfare of workers and firms? Put another way, how do wage floors affect producer surplus and consumer surplus? Remember that in the context of labor, workers are the producers and firms are the consumers.

Let’s start by looking at the producers. Those workers able to find employment at the new wage and equal in hours to what they had at the previous wage are better off because they receive higher compensation for the same work. Workers who are unable to find work at the new wage and who had work at the old wage are worse off; they want to work but cannot and so have no compensation. Workers who did not work prior to the minimum wage and continue not to work after it see no change in their welfare.

Producer surplus for those that have work rises. But remember the caveat from chapter 11: this assumes that we are employing only those on the lowest part of the supply curve. There is no mechanism to ensure this, so the producer surplus shown in figure 12.11 is really a maximum possible surplus, and it is likely to be substantially lower.

On the firm side, consumer surplus falls. This occurs because firms must pay more for the workers they employ, and they employ fewer workers. Overall, societal welfare falls unequivocally, as seen in figure 12.11. The new wage creates deadweight loss, signifying a lower level of total surplus than prior to the wage floor due to the lower overall level of employment.

Studies that have looked at the impact of minimum wage laws have found different results. The broad picture appears to be that minimum wages do have a disemployment effect but that the effect tends to be quite small. The likely reason for this is the relative inelasticity of labor supply and demand curves at the bottom edge of the wage scale.

Inelastic labor demand and supply mean that the labor demanded and labor supplied are very insensitive to changes in wages. Figure 12.12 shows this inelasticity as very steeply sloped labor demand and supply curves. The graph shows that as wage is raised to a minimum above the competitive wage, there is very little impact on labor demanded and labor supplied. As a result, the difference between the two—that is, unemployment at the minimum wage—is relatively small, as is the deadweight loss.

Figure 12.12 Minimum wages with inelastic labor demand and supply

A final question we might ask is whether minimum wages are effective poverty-fighting tools. In this case, the evidence is more mixed. In the United States, relatively few of those working for minimum wages are key income earners for poor households. Many of these jobs are held by teens and young adults who work part time for some extra income and are from wealthier households. So while there seems to be a fairly limited detrimental impact associated with minimum wages, economists tend to prefer more targeted policies for fighting poverty, like the earned income tax credit, which transfers money directly to the poorest working households.

12.5 POLICY EXAMPLE
SHOULD THE GOVERNMENT PROHIBIT SELF-SERVICE GAS?

Learning Objective 12.5: Apply a comparative statics analysis to evaluate government prohibitions of self-service gas on labor markets and the market for retail gas.

The prohibition on self-service gas in New Jersey and Oregon is, in essence, a mandate that retail gas stations employ workers who perform the task of pumping gas into cars. In the labor market for gas station workers, this policy has the effect of increasing the demand for labor. In fact, such a policy has the double effect of increasing both employment and wages, as shown in figure 12.13.

Figure 12.13 Effect on wages and labor hours due to a ban on self-service gas

To analyze our policy question, let’s suppose that a state that does not currently have a ban on self-service gas imposes one. The resulting change in the demand for labor is the blue curve in figure 12.13.

Here we see that employment in the form of labor hours increases from $L_1$ to $L_2$ and wages increase from $w_1$ to $w_2$. If this was the end of the story, we might be tempted to conclude that the ban on self-service gasoline is unequivocally good.

However, we now have enough economic analysis tools to understand that labor is an input in the production of retail gas and that when the cost of this input increases, the supply curve of retail gas shifts up, as figure 12.14 shows. The increase in the cost of the labor input is due not only to the increased wage but also to the increased amount of labor required to pump gas. The increased cost of supplying a gallon of gas will cause retailers to supply less, shifting the supply curve leftward from $S$ to $S^{\text{Ban}}$.

Figure 12.14 Effect on retail gas supply due to a ban on self-service gas

The leftward supply shift has the effect of increasing the price of gas to consumers and decreasing the amount they purchase, leading to deadweight loss. The policy question becomes, “How big is this deadweight loss to society compared to the benefit to society of the increased employment among gas station attendants?”

We can gain some insight into this question by comparing the states of Oregon and Washington. Oregon prohibits self-service gas, while its neighbor, Washington, does not. According to the US Energy Administration, in 2010, Washington had roughly twice the population of Oregon but almost identical gas consumption per capita of four hundred gallons per year. Additionally, Oregon employed roughly 3.5 more people per gas station than did Washington, and there were fewer stations per capita in Oregon. Extrapolating from the difference in employment, economists have estimated that the additional cost from the ban adds roughly five cents to the price of a gallon of gas in Oregon compared to Washington. Given that the number of extra attendants in Oregon appears to be in the vicinity of three thousand and the population of Oregon is about four million, it is likely that the deadweight loss that comes from the higher price of retail gasoline is quite large relative to the societal benefit of the relatively few added employees.

https://cutt.ly/96Tpm8J