← Back to the lecture: Factor Markets and Labor, Part B: Coalitions and Constraints
Lecture 18 Recap
What a minimum wage does when many employers compete and when one employer does all the hiring, why a line of applicants makes prejudice cheaper, and how licenses and unions decide who gets in.
Class meeting 24 · November 17, 2026 · 10-minute review
Rules about wages and entry change who gets hired and on what terms. A minimum wage above the market wage raises pay for the jobs that remain, but when many employers compete it leaves fewer jobs and a line of people who want them. When one employer does all the hiring, a floor set where supply meets demand can raise both pay and jobs, but a floor set too high costs jobs again. A line of applicants makes prejudice cheaper for an owner. Licenses and unions both raise pay by limiting who can sell the work. For each rule, ask who gains and who is kept out.
How to use this page
This recap covers the November 17 class. Class first finished Lecture 17 (human capital and harder jobs), then covered all of Lecture 18: Florida’s $15 minimum wage, one town’s labor market with many employers and with one employer, training, prejudice and a line of applicants, a two-employer example of discrimination, occupational licensing, and unions. All five TopHat questions ran. Read the five-step diagnostic first, then the lecture sections, then the checks. The separate Lecture 18 Transcript has the full explanation.
Use these steps whenever a question puts a rule on a labor market: a wage floor, a license, or a union.
What you should be able to do
Class started with the plan for the rest of the semester (see below) and a short review of Lecture 17: a firm hires workers until the value of what the last worker adds equals the wage. Then it finished Lecture 17’s last two slides and moved to Lecture 18, which uses that hiring rule to ask what rules about wages and entry do to who gets hired.
End of Lecture 17 · TopHat
Skills from school, training, and experience are human capital. Like a factory, they take time and resources to build, and they pay off in the future. Unlike a machine, they cannot be taken from you: you take them wherever you go. They can wear out, through illness, time away, or new technology, and rules and prejudice can block what they earn. A lender cannot repossess your skills, so school can be harder to borrow for than a truck. A doctor earns more than an executive assistant partly because becoming a doctor takes far more human capital.
The instructor added a second view of college (not on the slides). On the signaling view, a degree pays not because it makes you more productive but because it shows employers who can finish a hard task over four years. He joked that questions like “is this on the exam?” push him toward the signaling view, but said he thinks both are at work: economics, for example, teaches ways of thinking that help in life and work. Pay that differs by major can fit either view.
If two jobs pay the same, people pick the safer or more pleasant one, so a dangerous or unpleasant job must pay more to attract workers. The slide’s example: deckhands can work on day boats, home each night, or on boats out for weeks. Both add the same fish, so demand is the same, but fewer will spend weeks away at any wage, so supply is smaller there. Day boats pay $80 a day and weeks at sea pay $100. The $20 is a compensating differential. At those wages no deckhand gains by switching. The instructor’s other examples (not on the slides): lumberjacks, crab boats, and garbage collectors, who are well paid for the skill the job needs; and his own job, which, with long breaks and free choice of research, likely pays less than he could earn doing data work at a large tech firm.
The pay gap between men and women (not on the slides). The instructor said the often cited “77 cents on the dollar” compares the median pay of all men with that of all women, who on average work in different jobs; men, for example, are more likely to take dangerous jobs. In his reading of the evidence, most of the gap goes away when like is compared with like. He also argued that if equally productive women were paid less, firms would gain by hiring them, which would push their pay up. The slides add the cautions: a pay gap alone does not show its cause, and U.S. law bars covered employers from discriminating in pay and hiring.
TopHat check (Lecture 17’s last question). Any deckhand can become a licensed captain through a short, cheap course that most pass, yet captains keep earning much more. Which explanation fits best? People must be paid extra to take charge of a boat and its crew. Cheap training removes one reason for a gap, but the jobs still differ: a captain answers for the boat and everyone on it. “More deserving” is not something economics can judge; “the gap cannot last” ignores that it does last; and deckhands do add to the catch (with no one to bait and haul the pots, a crab boat catches nothing). The instructor added that years of experience are also needed to become a captain, and no one would invest that time without expecting higher pay.
A minimum wage with many employers · TopHat
On September 30, 2026, Florida’s minimum wage rose to $15 an hour, the last step of a plan voters approved in 2020: $10 in 2021, then $1 more each year. From January 1, 2028, it will be adjusted each year for inflation (the instructor called that a bad idea). A minimum wage is a legal floor: covered workers must be paid at least that much. Under Lecture 17’s rule, no firm keeps a worker who adds $10 an hour if it must pay $15: it would lose $5 every hour. Who gains and who loses depends on what the market looks like without the floor.
Suppose one town’s market for entry-level work. Demand is what one more worker adds to employers; supply is how many will work at each wage. Many employers compete, so none can set the wage. At $13 an hour, employers want 900 workers and 900 want the jobs. Employers gain ½ × 900 × $9 = $4,050 an hour above what they pay (their consumer surplus, from $22 down to $13), and workers gain the same above the least they would take (their producer surplus, from $13 down to $4). (A student asked where the $9 comes from: $22 − $13 for employers, and $13 − $4 for workers.)
A floor matters only if it is binding: set above the market wage. A $10 floor would change nothing here. The instructor noted that in Boca Raton most entry-level jobs likely pay more than $15 already, while in rural parts of Florida the floor is more likely to bind.
At $15, employers hire only the 700 who add at least $15, while 1,100 want jobs. So 200 jobs are gone (the instructor called this the disemployment effect), and 400 more people want jobs than get them: the 200 who lost jobs plus 200 drawn in by the higher pay. That is a surplus of labor, and the law keeps the wage from falling to clear it. The 700 who keep their jobs get $2 more an hour: $2 × 700 = $1,400 an hour, paid by employers. The 200 lost jobs were worth $15 down to $13 to employers, and workers needed only $11 up to $13; those gains of ½ × 200 × $4 = $400 an hour no longer happen (deadweight loss; the slide assumes the jobs go to those who would take the least). The worker laid off now earns nothing. So, the instructor said, in a competitive labor market a minimum wage is not an obvious way to help low-skilled workers.
TopHat check. Employers want 900 workers at $13, 700 at $15, and 500 at $17; 900, 1,100, and 1,300 people want the jobs at those wages. With a $17 floor, how many are hired, and how many more want jobs than get them? 500 hired; 800 more want jobs than get them (1,300 − 500). Employers decide how many are hired when the wage is above $13.
Suppose a beginner adds $11 an hour while learning, and $16 once trained. With no floor, a firm can hire him at a low starting wage and train him: the training is part of his pay, his full wage. Most important training, the instructor said, happens on the job. With a $15 floor, paying $15 for $11 of work loses $4 an hour, and once trained he can leave for another employer, so the firm cannot win that back. Employers will want workers who are already trained, so he must learn first, with his own time and money. The floor raises the money wage, but training is pay too, and it can fall. So even workers who get the higher wage are not sure to be better off.
One employer · TopHat
A monopoly is a single seller; a monopsony is a single buyer, here of labor. Suppose one company does all the town’s hiring, and it must pay everyone the same wage. To hire one more worker it must raise pay for all its workers, just as Lecture 14’s airline had to cut the fare on every seat to sell one more. So the cost of one more worker rises faster than supply (the dashed line). From 600 to 700 workers, pay rises from $10 to $11 for all, and the wage bill goes from $6,000 to $7,700 an hour: $17 per added worker, while they add only $15.50 each on average. So the company stops at 600 workers and pays $10, though the last worker adds $16. Compared with many employers, it hires fewer and pays less, and there is a deadweight loss.
The instructor noted that a company that could pay each worker a different wage (wage discrimination, like Lecture 14’s price discrimination) would not face this problem; the slide assumes one wage for all. A one-company town is the usual example, though workers can always move.
Suppose lawmakers know the curves and set a floor at $13. Now the company pays $13 to everyone anyway, and 900 will work for $13, so one more worker costs just $13. It hires while a worker adds at least $13: 900, as with many employers, and pay rises from $10 to $13. No gains are lost. Here the floor makes hiring cheaper at the margin, so the company hires more.
Who is likely to face one employer? (not on the slides). In the instructor’s view, low-skill labor markets are very competitive: a worker who can run a cash register at McDonald’s can also do it at Starbucks, Taco Bell, or the campus bookstore, and would take Home Depot’s $15 over McDonald’s $10 for the same job. Buyer power is more likely for highly skilled workers whose skills fit few employers: NFL quarterbacks have 32 possible employers, who could collude (a buyers’ cartel, an oligopsony); CEOs of large companies may be another case. He noted that NFL players answered with a union, the NFL Players Association: market power on one side met with market power on the other.
Now set the floor at $15 in the one-employer town. Each worker costs $15, and the company hires while a worker adds at least $15: 700. That is 100 more than with no floor, but 200 fewer than at $13, and the deadweight loss is back. Why set a floor above $13? Lawmakers cannot see these curves; the curves differ from town to town and change over time; and one statewide number must cover markets where it is right, too high, or does not bind. The instructor added that politicians gain by telling voters they set the wage as high as they could. A raise is easier to see than a job never offered.
TopHat check. Suppose the town’s entry-level jobs all come from one company, which hires 600 at $10 with no floor. Lawmakers set the floor at $18. At $18, 1,400 people want the jobs, and 400 workers would each add at least $18. How many does the company hire? 400, fewer than with no floor. At $18 one more worker costs $18, so it hires only those who add at least $18. A floor that could fix a lone employer’s power helps only if it lands near the right number.
Prejudice and discrimination · TopHat
At $13 the town’s market clears: no one is waiting for a job. An owner who refuses equally productive workers out of prejudice must find others; if that means paying more or leaving jobs empty, the prejudice cuts his profit. At $15, 1,100 people want 700 jobs. An owner can turn an applicant away and hire the next one, just as good, at the same $15: the prejudice costs him no wage and no output. (Legal and reputation costs can remain.) So, the instructor said, a minimum wage in a competitive market makes discrimination cheaper, which anyone who thinks discrimination is wrong should count against it. The slide’s cautions: this is one force in a model, not a verdict on any wage law or on anti-discrimination law; prejudiced customers or coworkers, a lone employer, and poor information can keep prejudice alive; U.S. law bars covered employers from discriminating in pay and hiring on protected grounds; and a pay gap alone does not show its cause.
TopHat check. The town keeps its $15 minimum wage, but new businesses open and hire most of the people waiting, so the line of qualified applicants is gone. Holding the law and reputation fixed, what happens to the cost to an owner of turning away a qualified applicant out of prejudice? It rises, because the owner may not find another worker as good at $15. Most of the class got it. More competition for workers makes discrimination more costly.
Suppose 2,000 equally productive workers: 1,200 men and 800 women. Two groups of employers, A and B, have the same demand. With no prejudice, workers sort until both pay the same: each hires 1,000 at $10. (If one paid $11 and the other $9, no one would work for $9; workers would move until the wages matched.) Now A refuses to hire women and fires its 400. A is left with 600 workers, and its wage rises to $12. The 400 women go to B, which now has 1,400 workers, and its wage falls to $8.
Is that the end? No. A man working at B for $8 can earn $12 at A, so men move from B to A. They keep moving until both pay $10 again. A hires 1,000 men; B hires 200 men and 800 women. Prejudice changed who works where, not the wage. A gap can open only if the prejudiced employers offer more than 60% of the jobs: then B is too small to hire all 800 women at $10. The instructor’s point: if employers discriminate, supply and demand predict that it shows up first in who works where. Equal pay alone does not show there is no prejudice.
Licenses and unions · TopHat
An occupational license is a legal permit to do a job for pay; without it, you may not do the job. Students named dentists and therapists; the instructor added barbers, morticians, and vets. Certification is different: a credential, with no ban on others. Suppose a town’s hair braiders must get a license: fees, exams, training hours. Getting in costs more, so supply shifts left. Appointments fall from 50 to 40 a week, and the price rises from $70 to $80: buyers pay $10 more on 40 appointments, or $400 a week, and the lost trades are worth ½ × 10 × $20 = $100 a week. (The instructor walked through these numbers using therapists, his wife’s profession, which requires yearly license fees and classes.)
The case for a license is information: buyers may not be able to tell a good provider from a bad one before they buy. If the license tells buyers something real, it can help them. The instructor doubted this case in general (his example: the people who mount your tires could cause a deadly crash, yet need no license), but his main point holds either way: a license raises the price and lowers the amount sold. Whether that is worth it is a question to ask trade by trade.
TopHat check. A state long required 1,000 hours of training, paid for by the trainee, before anyone could braid hair for pay. It drops the rule and offers an optional certificate instead. Assume the old rule only raised the cost of getting in. What does supply and demand predict? More people braid and the price falls; customers can still choose a certified braider. Dropping the rule makes it cheaper to start braiding, so supply rises.
A union takes a competitive labor market and makes it a monopoly on the selling side. Suppose one union sets the wage for all of the town’s workers, and many employers hire. Like a monopoly, it raises the price by selling less. For one more job it must accept a lower wage for every member, so the gain from one more job falls faster than demand (the dashed line, like Lecture 14’s marginal revenue). From 600 to 700 jobs, the wage falls from $16 to $15, and total pay rises from $9,600 to $10,500: $9 per added job, while those workers need $10 to $11. So it stops at 600 jobs at $16, against 900 at $13 with no union. Members gain $5,400 an hour, employers keep $1,800, and $900 an hour of gains is lost: workers who would work for $10 are left out. (This assumes many employers, not a monopsony.)
At $16, 1,200 people want the 600 jobs. A worker left out who needs less than $14 gains by offering to work for $14, and an employer gains by hiring him, since he adds more than $14. If undercutting spreads, the union wage breaks down: back to $13, with 900 hired. It is the cartel problem from Lecture 14, and the instructor compared it to a group project: everyone wants the others to keep the deal while they cheat. So a union must keep rival labor out and stop undercutting: through labor law, strikes and picket lines, and rules on who may join. Federal labor law does part of that job: a union chosen by a majority of workers bargains for everyone in the unit, and the employer must bargain with it.
Announcements (as given in class)
This is the last week of new material. The November 19 class is the last one with new content. There is no class on Tuesday of Thanksgiving week (November 24). Two weeks from November 17 (Tuesday, December 1) is the exam review, and two weeks from Thursday (Thursday, December 3) is the final exam.
The final exam is cumulative, with 30 questions instead of 40, so you have more time per question than on the midterms: 10 on the material before the first exam, 10 on the material between the first and second exams, and 10 on the material since the second exam. The questions will be like the ones you have seen, not the same questions.
Study guide and practice exam. The instructor plans to post a study guide and one practice exam on Thursday night, November 19, two weeks before the final. He noted that fewer than 10% of the class has been taking the practice exams.
Do not memorize an example as a story. Use it to recover the economic principle.
| Example from class | Economic lesson |
|---|---|
| Day boats $80 a day, weeks at sea $100 | A harder or less pleasant job must pay more to attract workers: a compensating differential. |
| The town: 900 hired at $13 with no floor | Many competing employers: the wage settles where supply meets demand, and each side gains. |
| A $15 floor: 700 hired, 1,100 want jobs | A binding floor with many employers cuts jobs, creates a line, moves pay to those who keep jobs, and loses gains. |
| The beginner who adds $11, then $16 once trained | Training is part of pay; a floor can push its cost onto the worker. |
| One company: 600 hired at $10, though the last adds $16 | A lone employer must raise everyone’s pay to hire one more, so it hires fewer and pays less. |
| A $13 floor gives 900; $15 gives 700; $18 gives 400 | A floor can undo a lone employer’s power only if it lands near where supply meets demand. |
| NFL quarterbacks and their 32 employers | Buyer power is likelier where skills fit few employers than in low-skill jobs. |
| 1,100 applicants for 700 jobs at $15 | A line of equally good applicants makes prejudice cheaper for an owner. |
| A refuses women; men move from B to A; $10 at both | With enough employers who do not share it, prejudice sorts workers instead of lowering their pay. |
| Braiding licenses: $70 to $80, 50 to 40 appointments | A license raises the cost of getting in: supply falls, price rises, trades are lost. |
| A union at $16 with 600 jobs; undercutting at $14 | A union raises pay like a monopoly seller and, like a cartel, must stop its own members and outsiders from undercutting. |
These use new numbers. Questions 2 to 8 and 11 use one example: a county’s market for entry-level warehouse work. Employers want 400 workers at $24 an hour, 600 at $20, 800 at $16, and 1,000 at $12 (demand falls by 200 workers for each $4). People who want the jobs: 600 at $14, 800 at $16, 1,000 at $18, and 1,200 at $20 (supply rises by 100 workers for each $1). Assume the jobs go to those who would take the least. Try each question before you open the answer.
Most people would rather not work at night, so fewer offer to work nights at any wage. The night shift must pay more to fill: the $2 is a compensating differential. At these wages no worker gains by switching. If many more workers came to prefer nights, the supply of night workers would rise and the gap would shrink. The gap is not a reward for merit; it pays for what the job asks.
Employers want 800 at $16 and 800 want the jobs, so the wage is $16 with 800 hired. Demand reaches $32 at zero workers and supply starts at $8, so employers gain ½ × 800 × ($32 − $16) = $6,400 an hour, and workers gain ½ × 800 × ($16 − $8) = $3,200 an hour.
At $20 employers hire 600 and 1,200 want jobs. So 200 jobs are gone (800 − 600), and 600 more want jobs than get them (1,200 − 600). The 600 who keep their jobs gain $4 an hour each: $4 × 600 = $2,400 an hour, paid by employers. The 200 lost jobs were worth $20 down to $16 to employers, and workers needed $14 up to $16, so ½ × 200 × ($20 − $14) = $600 an hour of gains no longer happen.
Paying $12 for $9 of work loses the firm $3 an hour while the worker learns, and once trained the worker can leave for another employer. So the firm will want workers who are already trained. The worker must pay for the training first, with time and money, or is not hired. The money wage is higher, but the full wage, which includes training on the job, can fall.
To hire 600 it must pay the $14 that 600 workers need; to hire 700 it must pay $15 to all. The wage bill rises from 600 × $14 = $8,400 to 700 × $15 = $10,500: $21 per added worker, while those workers add $20 down to $18, or $19 on average. Not worth it. One step back, from 500 at $13 ($6,500) to 600, the added workers cost $19 each and add $21 on average: worth it. So it hires 600 at $14, though the last worker adds $20. Compared with many employers (800 at $16), it hires fewer and pays less.
With a floor, one more worker costs just the floor (as long as enough people will work for it), so the company hires while a worker adds at least the floor. At $16: 800, the many-employer result, and no gains are lost. At $18: 700, more than with no floor but fewer than at $16. At $22: 500, fewer than the 600 it hires with no floor. A floor helps only if it lands near where supply meets demand.
(1) Knowledge: lawmakers cannot see the curves; they differ by county and change over time, and one statewide number must cover them all. (2) What voters see: a raise for workers who keep their jobs is easy to see; a job never offered, or training not given, is hard to see. Politicians can gain by saying they set the wage as high as they could. (3) Interests: groups whose members already earn more than the floor can gain when the low-wage labor that competes with them costs more. (Round numbers also campaign well.)
At $20, 1,200 people want 600 jobs, so the owner can turn an applicant away and hire the next one, just as good, at the same $20. The prejudice costs him no wage and no output (legal and reputation costs aside). With no floor the market clears at $16 and no one is waiting, so turning a qualified worker away can mean an empty job or paying more to fill it: the prejudice costs him profit.
Men at B can earn $18 at A, so men move from B to A until both pay $15 again. A hires 1,200 men; B hires 600 men and 600 women. Prejudice changes who works where, not the wage. A gap could last only if the prejudiced employers offered more than 75% of the jobs, since women are 25% of the workers: then B would be too small to hire all 600 women at $15.
Before: 100 − Q = 10 + Q gives 45 appointments at $55. After: 100 − Q = 20 + Q gives 40 at $60. Buyers pay $5 more on 40 appointments: $200 a week. The 5 lost appointments were worth $60 down to $55 to buyers and cost $50 up to $55 to supply, so ½ × 5 × $10 = $25 a week of gains is lost. If the license tells buyers something real about quality, that benefit must be weighed against these costs; certification could give the information without the ban.
At $20 employers hire 600, and 1,200 want the jobs. With 600 hired, one more worker adds almost $20, more than $18, so an employer gains by hiring him (at $18 employers would hire 700). The worker gains $18 − $15 = $3 an hour over the least he would take, instead of no job. If such offers spread, the wage slides back to $16, with 800 hired. Like a cartel, the union must keep rival labor out and stop undercutting: through labor law, strikes, and rules on who may join.
Bottom line
Start every labor-market rule from the market with no rule, then ask who sets the wage. With many employers, a binding minimum wage raises pay for the jobs that remain but cuts jobs, leaves a line, can push training onto workers, and makes prejudice cheaper. With one employer, a floor where supply meets demand can raise both pay and jobs, but lawmakers cannot see that point and politics has reasons to overshoot it. Prejudice by some employers tends to sort workers among employers rather than split their pay, as long as enough employers do not share it. Licenses and unions raise pay by limiting who may sell the work. For each rule, ask whom it lets in at higher pay, and whom it keeps out.
← Back to the lecture: Factor Markets and Labor, Part B: Coalitions and Constraints