Chapter 2 tested the belief in individual value at the top of the ladder. This chapter tests it at the bottom, where the same belief tells us that some work is worth very little.
Most companies do not decide what a role is worth. They look it up. A salary survey reports what other companies pay for the role, and the company pays about that. The figure is called the market rate, and we have come to treat it as a measurement of the work.
What the market rate measures
A market rate is a price. It records how many people are available to do a job and how much power they have to ask for more. Neither of those is the value of the work to the company, and three kinds of evidence show the difference.
The same work pays less under a different employer. When cleaning, security or catering is moved from a company’s payroll to a contractor, the work is unchanged and the pay falls. Chapter 13 cites the studies. If the rate measured the work, it would not depend on whose name is on the pay stub.
Pay follows who does the job. A study of American occupations across fifty years of census data found that occupations with more women in them paid less, after allowing for education and skill. The evidence pointed to the work being valued less because women did it.1
The work nobody can do without is among the lowest paid. In 2020, governments drew up lists of the workers who had to keep going when everything else stopped. In the United States, 22.3 million people on those lists worked in occupations with a median wage below $15 an hour.2 For a few months we were shown, very clearly, which work a society cannot go a week without. The market rate had been pointing the other way all along.
Vacancy Cost
There is a better question to ask of a role than what others pay for it. What happens if the work is not done? This paper calls the answer Vacancy Cost.
Take a software company with a hundred engineers and one maintenance worker, and give each of them a day off sick.
The engineer’s day. One engineer in a hundred stays home. The others pick up a review and move a deadline by a day. Nothing happens that a customer would notice.
The maintenance worker’s day. The maintenance worker has a full day of their own work. They are also the one person who knows the building and is there when something in it fails. Today something does.
- Overnight, leaves block a drain on the roof. By mid-morning, water from the cooling plant has backed up into an electrical conduit and tripped a breaker.
- The breaker feeds the network. A hundred engineers lose their connection. They cannot ship code or reach their tools, and the customer demonstration booked for eleven does not happen.
- On any other day this is a ten-minute problem. The maintenance worker knows the panel, has the key, resets the breaker and clears the drain. Today nobody knows which panel it is, or has the key to the plant room.
- A contractor arrives two hours later, finds the drain, and has the building running again after lunch.
At the average salary in the example company of Part Two, a hundred people idle for half a day is $25,000 of paid time, before counting the lost demonstration. That is half a year’s salary in that company’s lowest band, gone in one morning.
Two points in the story matter for later.
- Being available is worth something on the days nothing happens. What the company lacked that morning was a person who knew the building and was there. On most days that is never drawn on, which is why it is so easy to value at nothing. Sometimes the whole of a duty is to be available, and that is still work the company depends on. A duty sits beside a role, though, and is never the whole of one.
- The company relied on one person for it. Nobody else could find the breaker. Chapter 13 treats availability as a duty in its own right, with training and pay attached, shared among several people and backed by a supplier.
What Vacancy Cost does not do
It would be easy to take the wrong lesson from the story and conclude that the maintenance worker should be paid more than the engineers. Vacancy Cost cannot be used as a price, for two reasons.
- It measures the staffing. The engineer’s absence cost little because ninety-nine others could cover. Had the company employed two maintenance workers, the second would have been there. The cost of a vacancy tells us how thinly a function is staffed. It tells us nothing about the person.
- It has no ceiling. Take away all the engineers and there is no product. Take away all the maintenance and there is no building. Each function, removed entirely, costs the company everything.
What Vacancy Cost shows is narrower and more useful. The ranking in the salary survey is not a ranking of importance. Every role a company needs regularly is necessary to what the company produces, and “necessary” does not come in amounts that can be priced. This is Chapter 1’s argument arriving from the other end. The product is made together, and no method divides it among the people who made it.
Essential Employment
FairShare draws a principle from this and calls it Essential Employment. If a company needs a piece of work done regularly, the person doing it is part of the company, and is paid by the same rules as everyone else in it. No function is priced separately.
In practice that means one salary ladder for every function. A person’s place on it depends on their mastery of their own work, which can be seen, as Chapter 1 said. It does not depend on what a survey reports for their trade. An experienced maintenance worker and an experienced engineer who work at the same level are in the same band. Chapter 8 sets out how a band is defined so that this is possible.
It also means a company cannot keep someone on site every day while paying them by a different logic through a contractor. Chapter 13 states the test: if the work is needed regularly, the person is on the payroll and on the ladder.
Who the market rate costs
The market rate does not fall on everyone alike. It falls hardest on particular people.
- People in manual and service work. Their trades are priced low by the survey, whatever the work does for the company.
- Women. In 2024 American women earned 85 percent of what men earned, by median hourly pay.3 In 2022 Black women earned 70 percent of what White men earned, and Hispanic women 65 percent.4
- People who cannot safely bargain. Where pay is negotiated, it rewards those who ask. In experiments, evaluators penalized women more than men for asking for higher pay, and the penalty came mainly from male evaluators.5
One ladder with fixed, published figures removes two of the mechanisms behind these gaps. Pay is no longer set by the trade a person is in, and nobody negotiates. Chapter 8 cites evidence that publishing pay narrows the gap between men and women.
It does not remove bias. People still decide who is hired and which band a colleague belongs in, and people can be unfair. What FairShare does is make the result visible, because placements are published. Whether a company acts on what it sees is up to the people in it.
Notes
- Levanon, A., England, P., & Allison, P. (2009). Occupational Feminization and Pay: Assessing Causal Dynamics Using 1950–2000 U.S. Census Data. Social Forces, 88(2), 865–891.
- Kinder, M., & Stateler, L. (2021). Essential Workers Deserve a Minimum Wage Increase. Brookings Institution. https://www.brookings.edu/articles/essential-workers-deserve-minimum-wage-increase/
- Pew Research Center. (2025). Gender Pay Gap in U.S. Has Narrowed Slightly over 2 Decades. https://www.pewresearch.org/topic/gender-lgbtq/economics-work-gender/gender-pay-gap/
- Pew Research Center. (2023). The Enduring Grip of the Gender Pay Gap. https://www.pewresearch.org/social-trends/2023/03/01/the-enduring-grip-of-the-gender-pay-gap/
- Bowles, H. R., Babcock, L., & Lai, L. (2007). Social Incentives for Gender Differences in the Propensity to Initiate Negotiations: Sometimes It Does Hurt to Ask. Organizational Behavior and Human Decision Processes, 103(1), 84–103.