Why Openings Can Grow When Employment Shrinks
The BLS projections separate the size of an occupation from the flow of openings through it. This chapter follows that flow: how new positions, worker departures, replacements, and movement between occupational categories can produce many openings even when the ending employment level is flat or lower. It also separates the industry lens from the occupation lens, so a broad claim such as “healthcare is growing” can be translated into the more useful question of which kinds of work are expanding, replacing workers, or changing their requirements.
The central puzzle is easy to state. If an occupation has fewer jobs at the end of a projection period, how can it still offer opportunities during that period? The answer is that employment is a stock, while openings are a flow. A stock is counted at a point in time. A flow counts movement across time. The two measures interact, but they are not substitutes.
A job stock can shrink while the hiring flow continues
Imagine an occupation with 1,000 positions at the beginning of a decade. If 100 new positions are created but 140 workers leave positions that are then eliminated, the ending stock is 960. The occupation has contracted by 40 positions. Yet 100 people may still have been hired into newly created work, and some of the remaining positions may have been filled more than once as workers departed. The endpoint records the 40-position loss. An openings measure is designed to capture more of the activity that occurred on the way to that endpoint.
The example is deliberately simple. Actual projection systems do not follow every individual worker or employer. They estimate employment by occupation and then estimate the average opening volume associated with the projected labor market. But the stock-and-flow logic remains essential. A negative numeric change tells us that the occupation is expected to have a smaller employment base in the final year. It does not tell us that no employer will recruit during the decade, that every current worker will remain, or that an entrant has no possible route in.
Several transitions can contribute to the flow. An occupation may gain new positions because demand for its work expands. A worker may leave through retirement or another form of permanent separation, creating a replacement need even if the employer keeps the position. A person may move from one occupational category to another, creating an opening in the category they leave and changing the distribution of work elsewhere. The final employment level is the balance after these movements; the annual opening estimate reflects a broader stream of opportunity associated with them.
The relationship can be expressed without pretending that the data follow individual careers. Over a period, the change in the employment stock is roughly the positions added minus the positions removed. The opening flow is wider: it includes some of the positions that must be filled along the way, whether the final stock is larger or smaller. If ten positions are added and five existing positions are eliminated after their workers leave, the stock grows by five, but the period may still contain more than five hiring events. If the remaining positions are repeatedly refilled, the number of people passing through the occupation can be larger again. The endpoint is a balance; openings are activity along the route to that balance.
This is also why “replacement” should not be treated as a minor footnote. In a large occupation, even a modest rate of departures can create a substantial recurring flow. In a small occupation, a high growth rate may create only a small number of new positions and only a narrow replacement pool. A person looking for a first foothold may value the latter for its preparation and pay, while a training system may need to serve the former because the total number of entrants is much larger. The same labor market can look promising or limited depending on whether the decision concerns scale, access, or long-run expansion.
The sequence is not a literal timetable in which every event happens once and in that order. It is a map of the different kinds of movement that can coexist. New positions and replacements can happen in the same occupation. Workers can enter, leave, and move while the occupation’s total rises, falls, or barely changes. The diagram’s point is to keep the endpoint and the flow visible at the same time.
The office and administrative support group makes the distinction concrete in the current BLS table. Employment is projected to decline by 752.1 thousand jobs, or 4.0 percent, from 2025 to 2035. The same group has 1,727.7 thousand annual-average openings. A reader who sees only the decline may assume the work is disappearing immediately. A reader who sees only the openings may assume the occupation is expanding. Both conclusions are wrong. The table says that the long-run stock is expected to shrink while a substantial flow of positions still becomes available as the workforce changes.
Production has a similar relationship: a projected decline of 33.5 thousand jobs, or 0.4 percent, alongside 867.7 thousand annual-average openings. These numbers are not inconsistent because they describe different dimensions. The decline is the difference between two employment snapshots. The openings estimate describes expected opportunity volume over the intervening period. One is a direction-of-stock measure; the other is a flow measure.
Replacement demand is not automatically good or bad. It can create an accessible entry point for someone who needs work, but a high volume of replacements can also coexist with difficult conditions, low retention, or frequent movement out of the occupation. The available projections do not tell us why every opening occurs or whether every position is full-time, stable, local, or well paid. Openings indicate that work is likely to need people; they do not describe the entire experience of being one of those people.
The Georgetown Center on Education and the Workforce makes the boundary more precise in its 2021–31 analysis. Its annualized-openings measure counts new jobs and jobs created when workers move from one occupational category to another. It does not count job-to-job moves within the same occupational category. CEW also explains that its newer presentation uses annual average openings rather than a single cumulative total of new and replacement jobs, because it considers that a better representation of employment demand. Those choices affect the size and meaning of the reported number.
The choice of an annual average is useful for another reason: it keeps a decade-long opportunity claim from sounding like one giant pool of vacancies available at a single moment. An estimate of 800,000 openings per year does not mean 8 million jobs are simultaneously waiting to be filled, and it does not mean the same people will occupy all of them. Some positions will be filled and later reopened; some will be new; some will disappear as the occupation changes. The measure is best read as the expected pace of opportunity, not as an inventory.
CEW’s broader pattern is that large occupational clusters and lower-skilled, lower-wage occupations generally produce more annual openings than smaller, more selective groups. Occupations that employ many young workers, including retail sales and food and personal services, can produce large flows as people move on from early jobs, often while gaining more education. This does not mean every worker leaves, every job is temporary, or every high-opening occupation has poor prospects. It means that opening volume partly reflects the size and movement of the workforce, not simply the creation of new long-term positions.
The industry is not the occupation
The same stock-and-flow distinction becomes more useful when we change the unit of analysis. An industry groups organizations by the economic activity they perform. An occupation groups people by the work they perform. A hospital, home-care provider, or medical research company belongs to a healthcare-related industry, but each may employ nurses, aides, managers, technicians, analysts, food-service workers, and maintenance staff. Those occupations can have different growth rates, opening volumes, wages, and preparation requirements even when they share an employer or sector.
An industry projection is therefore closer to a statement about demand for an organization’s output. An occupation projection is closer to a statement about the kinds of tasks employers will need performed. The two statements are connected by the way organizations are staffed. When a hospital expands, the added demand may reach clinicians, aides, administrators, information specialists, and facilities workers in different proportions. When a technology-intensive industry expands, the effect may be strongest in a small set of technical occupations rather than evenly distributed across everyone employed by that industry. The staffing mix is the bridge between the two tables.
This is why an industry headline should be treated as a map of where demand is concentrated, not as a description of every job inside the sector. BLS projects private healthcare and social assistance to add 2,204,800 jobs from 2025 to 2035, growing 9.5 percent and accounting for about 37 percent of all new jobs. The release links that expansion to an aging population and the increasing prevalence of chronic health conditions. That is a strong sector-level signal. It does not say that every healthcare occupation will grow at 9.5 percent or that every new position will have the same entry route.
The occupation-level rows show the internal differences. Healthcare support occupations are projected to grow 13.3 percent, adding 1,122.7 thousand jobs and producing 1,262.8 thousand annual-average openings. Healthcare practitioners and technical occupations grow 8.0 percent, adding 819.5 thousand jobs and producing 605.0 thousand annual-average openings. BLS reports median annual wages of $38,340 for healthcare support and $86,530 for healthcare practitioners and technical occupations. The support group grows faster and has more openings, while the professional and technical group has a higher median wage and a different preparation profile.
At the detailed level, the contrast is sharper. Home health and personal care aides are projected to grow 18.1 percent, adding 847.3 thousand jobs and producing 760.5 thousand annual-average openings. The typical entry-level education listed is a high-school diploma or equivalent, and the 2025 median annual wage is $35,800. Nurse practitioners are projected to grow 41.0 percent, adding 137.8 thousand jobs and producing 29.4 thousand annual-average openings. Their typical entry-level education is a master’s degree, and the median annual wage is $132,300. The nurse-practitioner rate is much higher, but the occupation is much smaller and more selective. The aide occupation combines a large base, high numeric growth, and a large flow of openings, but not the same earnings outcome.
These differences also explain why healthcare is a better case than a simple list of fast-growing jobs. Aging and chronic disease can expand demand for care across a sector. That demand is then distributed among occupations with different tasks and constraints. A nurse practitioner, a home-care aide, a medical assistant, and a healthcare manager do not compete for the same role merely because their work is counted under a broad healthcare heading. The industry tells us where the pressure is. The occupation tells us how the pressure is translated into work.
The relationship works in the other direction too. One occupation can appear in many industries. Accountants, managers, drivers, software developers, and maintenance workers may work for organizations whose main activities are unrelated. Their occupational projections provide information about the work and typical preparation that travels across employers, while industry information adds context about the products, customers, regulation, or technology shaping that work. Neither lens fully replaces the other.
CEW gives private education services as a useful example of why industry totals can dilute education patterns. A private education organization may employ teachers, but it may also employ people in food service and building maintenance. If we ask how many jobs are in the industry, those roles belong in the total. If we ask what preparation is common for teaching, maintenance, or food service, we need the occupational lens. A sector’s average cannot be assigned to each component without losing the mechanism that created the average.
The distinction matters for decisions. A workforce program responding to healthcare growth might expand training for several occupations, but the correct programs depend on which occupations are actually growing and what entry conditions they have. A worker deciding whether to enter healthcare needs to compare the relevant occupation’s openings, education, wage, work experience, and local demand, not simply repeat the sector’s growth rate. An employer planning capacity may care about both: the industry signal identifies expanding demand, while occupational rows show where recruitment and replacement will occur.
The right interpretation of a large opening count is therefore conditional. It may mean that an occupation is expanding. It may mean that a large workforce is continually replenished. It may reflect both. To know which explanation matters for a particular decision, the reader must move through the layers in order: identify the industry or occupation, inspect its starting base, compare rate with numeric change, and then ask what the openings represent. Only after that does it make sense to examine preparation, wages, and the quality of the work.
Healthcare makes the sequence visible. The industry’s scale explains why a near-fastest growth rate becomes millions of new jobs. The occupational groups show why support work can grow faster and produce more openings than professional work. The detailed rows show why a very high rate can describe only a small number of openings, while a large opening flow may come with lower pay or demanding conditions. The next question is how these different routes into the same broad sector are shaped by education and training—and why the projection’s education fields must be read carefully.