"Baumol's cost disease" is a phenomenon that economists sometimes cite when they want to underscore the unequal effects of technological change.

William Baumol, the New York University economist for whom the concept is named, uses a classical string quartet as his paradigmatic example. As he points out, no matter how fast our computers become or how many people worldwide plug into broadband Internet access, it will still take roughly 30 minutes for four human beings to play Mozart's String Quartet No. 14.

If your business is playing string quartets for live audiences, there's a fundamental limit to how much more productive, in raw economic terms, your work can become.

Or, to take an example in which many more livelihoods are at stake, consider office cleaning. In the future, there might be new solvents, new cleaners, better scheduling software. But at the end of the day (usually literally), someone will still have to run a vacuum over a rug, run a rag over a table, empty smaller trash cans into bigger ones. Because of this, it's almost simple logic that the only ways for an office- cleaning company to make more money will always stay the same: Charge clients more or pay workers less.

In a big city, filled with competitive office-cleaning firms and a seemingly endless supply of desperate workers, this means office cleaners as a rule will never make much more than the minimum wage.

Which is why, just over a year ago, Mr Guillermo Garcia was perplexed by a job opening he heard about. He had been unemployed for six months - unable to find even minimum-wage work - when his counsellor at Madison Strategies Group, a job-training-and-placement firm for low-income people in New York City, told him about the office- cleaning jobs that a brand-new company had listed.

The firm was called Managed by Q, and it was paying US$12.50 (S$17.50) an hour, or more than 40 per cent above the city's minimum wage. Even more unbelievable, the job offered full healthcare benefits and a 401(k) retirement plan.

Mr Garcia, 24, was born in a small, poor village in Central Mexico. His parents brought him to New York City when he was nine. None of them has legal documentation. His father works at a carwash, his mother cleans houses. They don't make much. Mr Garcia finished high school in 2010 - something his parents never did - but graduated into an economy that had a brutally low income ceiling for an undocumented worker like him, even one who grew up in the United States.

The New York State Department of Labour recently released a report about salaries in New York City, and it makes for grim reading. Entry-level salaries for the kinds of jobs that someone like Mr Garcia might apply for - retail clerk, food preparer, restaurant server, janitor, unskilled construction worker - are stuck right at or barely above the minimum wage, which is now US$18,700 (S$26,200)a year.

A few weeks after hearing about the improbable job at Managed by Q, Mr Garcia had it. And today, just over a year later, his pay is up to US$14.50 an hour, and a remarkable US$21.75 when he works overtime, which he does as often as he can.

His supervisor is talking to him about a promotion that would come with a higher salary. There are even rumours of stock options. His life has been transformed, almost as if he won the lottery.

MANAGED BY Q

The chief executive of Managed by Q, Mr Dan Teran, wants to be clear: His company is not a charity. Its plan is to become enormous and highly profitable.

A little more than two years old, the firm now cleans more than 2.1 million sq ft of office space in New York City. (The fanciest two tiers of buildings, referred to as "trophy" and Class A in corporate-real-estate speak, typically hire their own in-house cleaning services, but Managed by Q now cleans more than 1 per cent of the rest, the so-called Class B and Class C offices.)

In addition to cleaning, the company offers a whole suite of other services to its clients - maintenance, IT support, security, supplies and others. It has expanded its operations to San Francisco, Chicago and Los Angeles, with the eventual goal of conquering every major city in the United States and, eventually, the world.

The company pays its staff, like Mr Garcia, considerably more than prevailing market rates not solely because its founders want to be kind to them, but because Mr Teran sees it as crucial to his business model.

He believes that most US businesses, and especially fast-growing start-ups like Uber, have mistaken short-term gains for long-term value, undercutting the share of revenue that flows to workers in a way that will perversely hurt their bottom line.

He believes, even more radically, that decades of rising inequality and stagnant wages in America are not an inevitable by-product of capitalism; instead, they come from a simple misunderstanding about how best to deploy workers and recognise the value they bring to a firm.

The future of jobs in the US would be very different if Mr Teran's ideas catch on. But first, of course, he has to prove that they actually work.

The concept behind Managed by Q came when Mr Teran and Mr Saman Rahmanian, friends who met through New York's tech scene, realised that the various start-ups they had worked for always experienced the same weird problem. The office itself - the physical place in which the start-up worked - could never adapt as quickly as the company needed it to. Why was it such a pain to set up a new work space, or move a wall, or order new desk chairs, or bring in an exterminator?

They imagined the dream office manager as something like the character Q in James Bond films, a quiet genius who understands technology and creates the tools that help Bond do his work. Thus the name: Managed by Q.

The business would first present itself to new clients as a cleaning company. That's because potential customers already know what a cleaning company is, and they're often looking for a new and better one. But once the clients were sold on the high quality of the cleaning service, they would come to realise that Q (as employees call it) could do so much more.

Every new client was given a wall-mounted iPad with an intuitive app. Someone could request through the iPad, say, more paper towels or an exterminator or furniture assembly or that a wall be put up or taken down. Q would take care of all the details, quietly and quickly.

At first, Mr Teran was worried that his new idea would force him to make a horrible choice. He wanted to create a company that could pay a solid living wage to its workers and offer them good benefits. But he couldn't figure out how to make the numbers add up. After seeking advice from anyone he could find, he found himself talking to Ms Dervala Hanley, who was then the vice-president for global strategy at Starbucks. She told him that the answers to all of his questions were in a book: The Good Jobs Strategy by Zeynep Ton, a professor of operations management at MIT.

MORE WORKERS, MORE PROFIT

The book offers a careful study of a handful of large companies that have been able to pay their workers well above prevailing wages while also keeping overall costs down and earning a healthy profit.

Typically, companies see workers entirely as a cost, something that reduces profit; as a result, many try to minimise the number of workers and the amount they are paid. But as Prof Ton points out, when a firm deploys its workers in smart ways, those workers can become a source of profit instead of just a cost.

Q is a perfect example: Mr Teran doesn't see his cleaners as simply cleaners. Instead, they are the essential link between Q and its clients.

Mr Teran told me that the best way to see this was by looking at a spreadsheet he had asked his chief financial officer to create. It was a simplified version of Q's own profit-and-loss statement alongside that of a fictitious competitor, based on standard industry data.

The spreadsheet assumes that Q and its competitor, Non-Q, each begin with 100 clients, meaning 100 offices to clean every day. Non-Q pays US$9 an hour for its workers; by hiring them as independent contractors rather than as salaried employees, Non-Q is also able to avoid paying Social Security taxes, overtime and other costs.

Q, by contrast, pays US$12.50 an hour plus benefits and Social Security taxes. This means that Non-Q realises a total profit, per office, of US$9,660, while Q makes a profit of only US$7,980. It seems, at first blush, that Q has a terrible business model. The more customers it has, the further Q will fall behind the competition.

But over time, the picture changes. The cleaning industry is famous for steep turnover in employees and customers; as with many commoditised, low-cost businesses, the end of every contract is an opportunity to see if someone else can do the job more cheaply. Industrywide, roughly half of a cleaning company's customers will choose a different firm every year, and it frequently costs about US$1,750 in marketing and sales to acquire a new customer. Also, roughly half of a cleaning company's employees will leave in any given year, and it costs around US$150 in advertising and human resources to refill each slot. Mr Teran believes that by paying workers more - and training them to stay in close contact with customers to make sure they are satisfied - he can drastically reduce both turnover numbers.

So far, it has worked: Only around 10 per cent of clients leave each year, and only 5 per cent of employees leave. This means Non-Q has to pay a lot of money every year just to keep the same number of customers and employees. Q can funnel that spending into expansion, acquiring more clients.

Mr Teran points out that on the spreadsheet, Non-Q does better in its first year of operation: It makes a net profit of US$871,000, compared with Q's US$709,450. It is only after a few years that the advantage of Q adds up. By the fifth year of operation, Q is bringing in 62 per cent more profit than Non-Q.

The real secret to Q's business, though, comes with the other services that it offers and facilitates through its app - maintenance, IT support, security and so on, some of which enjoy a significantly higher profit margin than its cleaning services do.

With these allied services, Q functions as a matchmaker between its clients and a marketplace of service providers, working with a few carefully screened firms and connecting its clients with them quickly; Q then receives a percentage of their fee. Already, 30 per cent of the company's revenue comes from services other than cleaning.

The whole model works only if the clients trust Q. And the main contact with Q is that cleaner, or operator, who goes to the office every day. One reason Q pays its cleaners so much is that they effectively become a sales rep, smuggled every day inside the office of the customer. The better the cleaner is - the better trained and motivated and paid - the more effective a representative she can be, and the more money she can generate.

For this reason, a main challenge - and one that will increase as the company grows - is selecting employees. Mr Teran has invested heavily in data technology that allows the company to study the correlation between specific operators and clients and assess what sorts of relationships work best.

Mr Teran and his team learnt that traditional markers of success - education, experience in the industry, recommendations from employers - were not tightly correlated with success at Q. Instead, the two crucial personal characteristics are optimism and empathy.

When asked what he meant, he suggested spending a shift with Ms Nancy Gonzalez, one of the most successful workers at Q. She is now a supervisor, which means she visits several different offices each night to check on the operators and offer help. She also works with the administrative staff, matching operators to clients. For example, one of her accounts is quite meticulous, she said, so she always makes sure the operator assigned to it is quiet and diligent; another account, by contrast, prefers a more engaging operator with a bigger personality.

Mr Teran remembers when Ms Gonzalez came in for interview; he was still able to meet all new recruits at the time. He gathered a group of recruits in a circle and asked them to describe a deeply negative experience. Ms Gonzalez talked about facing the collapse of the bakery that she and her boyfriend owned in the Bedford-Stuyvesant neighbourhood of Brooklyn. The goal was to see what kinds of words she used to describe the fallout of the negative experience - to see if she was able to think positively even when recalling the most painful time in her life. Ms Gonzalez's answer scored quite highly for optimism.

After that, she was invited to a group exercise to explore her ability to work well with others. Several potential employees were asked to describe embarrassing incidents from their past. As Mr Teran explained, the real goal of the exercise was to see how others behave when a person is presenting himself in a vulnerable way. Ms Gonzalez consistently showed enormous empathy, leaning forward, nodding along, unconsciously saying quiet, encouraging words. So she got the job, despite that fact that she had no cleaning experience, other than cleaning up her bakery and a home with seven children.

Her first assignment, when she joined in September 2014, was at the offices of a growing health-and- wellness company. The firm had a regular cleaning company that did a thorough job each night and used Q only for supplemental tidying during the day. The revenue from the firm was around US$400 a month.

The Q app allows customers to rate an operator's performance on any given day and, right away, Ms Gonzalez was receiving the highest grades every day - five stars, five stars, five stars. The client would put notes in the review about how Ms Gonzalez was a joy to be around, how she repeatedly went beyond the call of duty. Twice, she spent her cleaning shift filling in for a sick receptionist. Soon, the client replaced its night-time firm with Q. Then it asked Q to clean its other offices, as well as provide other services, like maintenance and supply management. Today, the client pays around US$7,000 a month to Q overall. So, in this one case, Ms Gonzalez was central to bringing in an additional US$79,200 in revenue per year.

Why don't more CEOs see the maths the way Mr Teran does?

In large part because all the short-term incentives of corporate America point them in the opposite direction. Suppose that in 10 years, Mr Teran's ambitions are realised and Q is a multibillion-dollar company with a million employees worldwide. By then, he might no longer be CEO and the company would be overseen by a board selected by its top investors. Just imagine the pressures on that chief executive.

If the company cut worker pay by US$1 a week, the firm would instantly realise a profit of US$52 million a year. Then imagine that the company cut wages by US$1 an hour. That would mean additional profit of US$2.1 billion. Maybe the chief executive would realise that cutting pay would inspire many to quit. But he could cut the working time in each office by, say, 10 per cent, allowing Q to lay off - or not hire - 100,000 workers worldwide.

This is how lousy jobs and stagnant wages arrive.

Each company makes a series of relatively small decisions, based on what competitors are doing and what customers and investors are demanding. Soon, a new, lower norm is established, and the pressures are on everybody else to follow. In standard economic models, it's simply assumed: Wages fall to the point where supply meets demand. This is why Mr Teran plans that by the time Q is publicly traded - if it ever is - he will have so fully built out the model that even the most hard-headed successor will continue to implement it. And it's why he is also working hard to persuade other chief executives to take his ideas seriously.

This writer sat in on a phone call Mr Teran had this month with Mr Colin Barceloux, chief executive of ConvoyNow, a service that sends IT workers to private homes to fix computer emergencies. Mr Barceloux had assumed that he would use the Uber model, allowing independent IT professionals to receive a message any time somebody wanted help. He explained to Mr Teran that his company is so new that he has no idea how much demand there will be in any given city at any given time, so he can't hire people and pay them a fixed wage.

But Mr Teran countered that Mr Barceloux was thinking about the issue in the wrong way. If his customers had a different IT person each time they used the app, and the providers didn't end up giving the same level of service, the customers would fall away.

To follow the Q strategy, Mr Barceloux might need to rethink his business: Focus on fewer cities, or even start with just one, and build up the model locally. Perhaps there were other related services that ConvoyNow could provide, to take advantage of its trained workers.

I followed up with Mr Barceloux a few weeks later. He was still convinced by Mr Teran's argument, and he hoped to shift his business to a model closer to Q's, in which the people serving customers are employees, not independent contractors. But he wouldn't be able to offer those employees full-time work, or maybe even predictable schedules. As he pointed out, the peculiarities of the cleaning business - that it's a service that clients contract for every day, on a regular basis - are crucial to why Managed by Q can offer truly good jobs.

Will the strategy really work for other industries, including his own? Mr Barceloux was not entirely sure, but he planned to spend this year trying to figure out if it can.

NEW YORK TIMES

• Adam Davidson is co-host of the podcast Surprisingly Awesome and a contributing writer for the New York Times magazine, where this article was first printed.