The Jobs That Grow Are Not Necessarily New Work
This series studies new work in the U.S. labor market: the emergence of occupational specialties and tasks that require expertise not previously in demand. The immediate question is why that is different from simply having more jobs in occupations that already exist. The distinction gives readers a more reliable way to interpret employment forecasts, automation claims, wage statistics, and the question of who benefits from economic change.
The difference matters because “more jobs” can describe several events at once. An existing occupation can expand. A familiar occupation can absorb new duties. A genuinely new specialty can appear inside an established occupational category. These events may all increase employment, but they do not create the same learning requirements, wage opportunities, or transition paths.
The number that answers only one question
Consider the Bureau of Labor Statistics page titled “Most New Jobs.” It ranks the 20 occupations with the highest projected numeric change in employment from 2025 to 2035. Home health and personal care aides lead the list, with 847,300 projected new jobs. Stockers and order fillers are next with 250,700, followed by fast-food and counter workers with 223,300. The same table includes registered nurses, software developers, data scientists, managers, electricians, and other familiar categories. It also reports median pay in 2025—for example, $35,800 for home health and personal care aides and $135,980 for software developers.
That is valuable information. It tells us where the largest additions to recognized occupational categories are expected. It gives a sense of scale and shows that projected growth is spread across care, logistics, food service, health, management, software, construction, transportation, and professional work. It does not tell us that these occupations have newly emerged, that their internal tasks are new, or that the workers who fill them must acquire a new kind of expertise.
A projection of more nurses may mean that more people will perform a longstanding occupation. A projection of more software developers may involve new tools and duties, but the occupation-level number cannot identify which ones. A growing category may contain a mixture of older tasks, redesigned tasks, and genuinely new specialties. The category is a useful container for counting workers; it is not a direct measurement of novelty.
This is the first distinction to keep in mind:
- Headcount growth asks how many more people may be employed.
- Occupational expansion asks which recognized categories may contain more workers.
- New work asks whether new tasks or specialties have appeared that require human expertise not previously demanded.
The three levels can overlap, but none can stand in for the others. A category can grow without becoming new. A new specialty can emerge inside a category whose total employment changes little. And a job projection can be accurate while leaving the most important question about expertise unanswered.
The research paper at the center of this series uses “new work” in that narrower sense. It studies novel occupational roles that emerge as technology and economic conditions change. The focus is not whether a job sounds modern, pays well, or is associated with a fashionable industry. The focus is whether work has taken a new form and whether that form creates demand for expertise that was previously scarce or unnecessary.
That definition also prevents an easy mistake in the other direction. New work is not simply a label for high-tech work. The study distinguishes technology-linked titles such as an artificial-intelligence specialist from titles that can arise through changing tastes, income, or demographics, such as a conference planner. A new specialty can be technologically driven, but technological change is only one source of new demand.
Why novelty matters when technology changes tasks
The distinction becomes more consequential when technology changes what people do. In a simple story, automation replaces workers and the labor market must somehow find an equal number of other jobs. That story treats work mainly as a quantity: if one task disappears, another task must appear to preserve employment. But task-based economics asks a more precise question: which tasks are performed by capital, and which still require labor?
Technology can make capital a direct substitute for labor on particular tasks. Mechanical telephone switching, for example, sharply reduced telephone-operator employment in the cities where it was adopted. Such changes are not explained away by saying that there is always more work somewhere. A machine can perform a task that people previously performed, and the local demand for that task can fall quickly.
New work is the counter-process. It expands the set of tasks for which human labor is demanded. New tasks may arise because technology creates a new service, because production becomes more specialized, or because changing demand makes a previously unnecessary activity worth performing. In that sense, new work can reinstate labor’s role even as automation erodes labor’s role in other tasks.
This is not a promise that every technological advance creates enough work to replace what it removes. It does not say that displaced workers can immediately move into new specialties, that the new work will be located in the same places, or that the gains will be shared evenly. It identifies a mechanism that a count of jobs cannot show: technology and economic change can simultaneously eliminate some tasks and create demand for new human capabilities.
The paper’s central claim is that the value of new work comes from expertise. A new specialty can require a capability, skill, form of human capital, or kind of know-how that was not previously needed in production. But expertise is not automatically valuable merely because it is difficult or unusual. For a market return to exist, the expertise must be necessary for a good or service that people value and scarce enough that employers cannot obtain it effortlessly.
Novelty can create that scarcity, at least for a time. When a type of work first appears, few workers have experience doing it, and institutions may not yet have standardized training for it. Employers may therefore pay more for the people who can perform it. As more workers learn the specialty and knowledge spreads, the scarcity can weaken. The occupation may remain important, but its initial premium can fade.
This gives us a better mental model of labor-market change. The question is not only whether employment rises after technology or demand shifts. We also need to ask whether new expertise is being formed, how quickly it can spread, and which workers have a realistic path to acquiring it.
The diagram separates a visible result from the less visible process beneath it. A headcount projection may be the starting point for inquiry, but it is not the end. To call a change new work, we need evidence about the specialty or task, the expertise it requires, and the way workers and employers respond.
A better reading rule
The practical consequence is a simple discipline: whenever a report says that the economy will create more jobs, ask what kind of claim it is making before drawing a conclusion.
First, what is being counted? A numeric projection, a change in an occupational category, a newly appearing title, a task, a wage, or a worker’s transition are different observations. The BLS list is designed for the first two. It should not be treated as a direct answer to the third.
Second, what is actually new? A new title can be evidence of an emerging specialty, but a title is still a proxy for the underlying tasks. Established occupations can contain new duties, and a newly recognized label can include familiar work. A serious interpretation must keep the classification rule in view.
Third, what expertise is scarce? If the change requires only more workers doing the same established tasks, the relevant issue may be recruitment capacity or working conditions rather than new-work expertise. If it requires knowledge that few workers currently possess, the likely questions become training, learning time, location, and the durability of the demand.
Fourth, who can reach the opportunity? New work may expand labor demand while favoring workers who are younger, more educated, urban, or already close to the relevant networks and training. A growing occupation is not automatically an open route for every displaced worker.
Finally, how will the return change? Early scarcity can produce a premium, but that premium may decline as skills diffuse, demand changes, or the specialty becomes ordinary. A median-pay figure describes the level of earnings in an occupation; it does not by itself identify the return to acquiring newly scarce expertise.
The rest of this series follows those questions through the evidence. It examines how researchers identify new work in occupational records, who performs it, why it can carry a wage premium, how that premium changes with time, and how a demand shock can create new specialties in particular places. The goal is not to replace job counts with a more fashionable label. It is to understand what job counts leave out.
The key conclusion for now is modest but important: more work is a quantity; new work is a change in the content and expertise of work. Both matter. Confusing them makes it difficult to interpret automation, to judge the value of training, or to see why employment growth can benefit some workers more than others. The next question is therefore operational: how can new work be observed rather than assumed from a rising number of jobs?