Who Gets to Enter the New Economy?
The question in this chapter is not simply how many new jobs exist, but which workers are most likely to be doing new work once it appears. The evidence points to a pattern: new-work participation is associated with education, age, and urban location, even when comparisons are made inside detailed occupations and places. That pattern helps explain who is positioned to acquire emerging expertise, while the study’s observational design cautions against treating any demographic trait as a simple cause of entry.
The distinction matters for the meaning of employment growth. If a growing occupation mainly needs more people to perform established tasks, access may depend on recruitment, working conditions, and the availability of workers. If it is absorbing new specialties, access may depend more heavily on prior education, proximity to specialized employers, timing, and the ability to learn unfamiliar skills. The same headcount increase can therefore create very different opportunities.
Education is linked to entry into new specialties
The IZA study finds that new work is disproportionately performed by more educated workers. In the contemporary data, the incidence of new work rises monotonically with educational attainment. The pattern remains when the researchers compare workers within detailed occupational categories and when they add industry and county controls. In the preferred specification, workers with advanced degrees are 2.9 percentage points more likely than high-school graduates to be in new work, against an overall contemporary new-work rate of 18.3%.
The historical pattern points in the same direction, even though the occupation mix was different. In 1950, workers with a bachelor’s degree or higher were 1.3 percentage points more likely than high-school graduates to be in new work, compared with a mean new-work rate of 7.2%.
These are not enormous gaps in absolute terms, but they are meaningful relative to the base rates. A 2.9-point difference is a substantial share of an 18.3% incidence. More importantly, the relationship survives increasingly demanding comparisons. It is not explained only by the fact that highly educated workers are concentrated in occupations that pay more or that happen to have more new titles. The pattern also appears within broad occupational and geographic settings.
Why might education be associated with new work? New specialties often require workers to combine existing knowledge in unfamiliar ways, interpret new tools, or learn concepts for which standard workplace routines have not yet formed. Formal education can supply general cognitive and technical foundations that make such learning easier. It can also act as a signal to employers who are deciding whom to trust with an emerging responsibility before performance histories exist.
Those are plausible mechanisms, not findings that the study directly proves. The data do not show that adding a particular credential mechanically produces entry into new work. Education can be correlated with family resources, networks, prior opportunities, employer selection, and unmeasured ability. The safest conclusion is narrower: workers with more education are more likely to be observed in new specialties, even after the researchers account for several features of their occupational and local context.
That distinction is important for how the result should be used. It does not justify telling every worker to obtain a degree in the hope of reaching an unspecified future occupation. It does show that a labor market generating new work may reward capabilities that are not yet visible in the title of the job. A report about job growth is therefore incomplete if it describes demand but says nothing about the learning path needed to meet it.
Age and place shape the opportunity to learn
Age produces another consistent pattern. Workers under 30 are most likely to be in new work, workers aged 55 to 64 are least likely, and those aged 30 to 54 fall between them. The result fits the idea that younger workers may find it easier to invest in skills for an emerging specialty or may face fewer costs when changing direction. It could also reflect the way employers allocate opportunities, the fact that younger cohorts entered the labor market after a specialty appeared, or the accumulation of experience in older forms of work.
Again, the age pattern is an association. It should not be read as evidence that older workers cannot learn new work. A worker who already has deep knowledge of an industry may be well placed to recognize and perform a new specialty, even if the average rate of entry is lower at older ages. The data describe who is more likely to be observed in the classified new titles; they do not establish a biological limit or a universal career rule.
Location matters as well. Conditional on education, age, occupation, industry, and county groups, urban workers are more likely to be in new job titles. The urban concentration is not merely a consequence of cities containing more high-paying occupations. The association persists even within new-work-intensive occupations and within counties.
A place can make emerging work easier to find and easier to learn. Specialized employers may be closer together. Workers may encounter more firms experimenting with new processes. Training providers, professional networks, and complementary services may appear around the same demand. These mechanisms are consistent with the study’s findings, but the regression results do not separately identify which one is responsible for the urban association.
The location result also changes the meaning of a national forecast. A projected increase in an occupation is not a map of where new expertise will form. A specialty may be concentrated in particular cities, industries, or firms before it spreads. A national occupation count can hide that geography, just as it can hide the difference between older duties and new ones.
The paper also reports conditional demographic differences beyond education, age, and place. Holding several characteristics fixed, white workers, women, non-self-employed workers, and urban workers are more likely to be employed in new job titles in both periods. These patterns are descriptive associations, not a ranking of who is inherently suited to new work. They may reflect unequal access to education, occupational sorting, employer decisions, local opportunity, or the particular composition of the titles included in the measure. The useful question is not which group is “best” at new work, but which institutional and labor-market conditions shape entry.
Why within-occupation comparisons change the interpretation
A simple occupational ranking can create a misleading picture. Suppose new titles are common among software developers, managers, or professional workers. If we compare all workers in those occupations with all workers elsewhere, the result may reflect occupation-level pay, education, or location rather than anything distinctive about new work. The study addresses this by comparing new and preexisting work within detailed occupational categories and, in some specifications, within industries and counties.
The within-category comparison does not make the result causal, but it changes the question. It asks whether two workers located in the same broad occupational setting differ in their likelihood of reporting a new title. The answer remains yes for education and several other characteristics. That suggests that new work is not merely a label for a handful of occupations that are already high in the earnings distribution.
The comparison also reveals why the phrase “new work” does not name one social class. In the historical data, more than four in ten workers in new titles were in production or clerical and administrative support. In the recent data, more than four in ten were in the two highest-paid broad categories, professional and managerial. New work has moved across the occupational distribution as technology, production, and consumer demand have changed.
Education changes the location of that shift. Among workers with at least four years of college, new work is heavily concentrated in professional and managerial occupations in both eras, with technical work a larger contemporary category. Among workers with less than four years of college, production and clerical support accounted for more than 40% of new work in 1940–50, but less than 25% in the contemporary data; health and personal services now account for more of their new-work employment.
This is a distributional fact with a practical implication. A labor market can continue to generate new specialties while changing the kinds of capability that lead to them. The opportunity may move from a production process that rewards shop-floor experience to a service or technical environment that rewards formal education, or it may move into care and personal services without offering the same pay or training path. “New work” identifies novelty; it does not guarantee upward mobility.
The finding also limits any simple story about technology. The paper’s technology-linked classification suggests that more than half of new titles are plausibly tied to technology, but the broad occupational pattern includes non-technological sources of change as well. Contemporary new work can emerge in professional and managerial settings, but it also appears in health and personal services. New work is shaped by the interaction of technology, demand, demographics, and institutions.
A growing occupation is not automatically an open door
The BLS projection table makes the distinction visible in a different way. It includes large projected increases in home health and personal care aides, stockers and order fillers, fast-food and counter workers, registered nurses, and software developers. The categories represent different kinds of work, but the table itself does not report their education, training, working conditions, or task content. It reports the expected number of additional jobs and median pay; it does not report who can move into the roles or whether the roles contain emerging specialties.
It would be a mistake to infer that the largest category is the most accessible new-work opportunity. A large number may reflect the size of an existing occupation, not a low barrier to entry. Conversely, a smaller category may require substantial expertise and generate important new tasks without producing enough openings to rank near the top. Median pay also cannot settle the question. It describes earnings within the recognized category, not the cost of acquiring a new specialty or the return to doing a newly defined task.
A more useful interpretation asks three separate questions. Does the occupation need more workers? That is the projection question. Are the tasks or specialties changing? That requires task or title evidence. Who can acquire and reach the relevant capability? That requires evidence about education, age, geography, training, and transitions. Treating these as one question hides the actual opportunity structure.
The diagram is intentionally modest. The observed variables are associated with participation; they do not form a proven pipeline in which one factor mechanically produces the next. Education may improve preparation, but employers and institutions still control opportunities. Urban location may expose workers to emerging demand, but not every city resident can reach the relevant employer or training. Age may correlate with timing and skill investment, but it does not determine an individual’s capacity to learn.
The evidence therefore supports a specific judgment. New work is not distributed randomly across the workforce, and the distribution has consequences for whether employment growth becomes shared opportunity. The strongest patterns point toward education, younger age, and urban context, including comparisons that hold occupation and place more constant. The explanation remains open enough to include skill acquisition, employer selection, networks, and local demand.
That is why access belongs in the definition of what employment growth means for workers. A national count can rise while the route into emerging expertise becomes more concentrated. Before asking whether new work pays more, we need to recognize that the ability to reach it is already uneven. The next question is whether that scarcity of expertise appears in earnings—and whether the premium survives after the work stops being new.