What Workers Can Do—and What Each Move Costs
For fossil-fuel workers who lose their jobs, the subject is not mobility in the abstract but the four ways sector and place can change together. Comparing those paths helps us judge what a move actually repairs: access to another occupation, access to another employer market, both, or neither.
A worker leaving a declining industry has two distinct adjustment margins. They can change sectors, moving from fossil-fuel employment into non-fossil work, and they can change local labor markets, leaving the commuting zone where they lived and worked. Those choices can occur together, but they solve different problems. A local sector switch may preserve housing and family ties while opening more occupations. A geographic move may create access to more vacancies while preserving the same narrow occupational identity. Treating both as one decision hides the costs that follow.
Four routes out of a local shock
The Census study follows workers for seven years after separation and classifies them using both employment and residence histories. In its definition, a worker is a sectoral switcher if they work in the non-fossil sector for more than two years during that period. A worker is a mover if they leave the commuting zone where the separation occurred and do not return. These are long-horizon categories. They are not labels for what a worker necessarily decided in the first week after a job ended, and they do not capture every temporary job, daily commute, or return move.
The combination produces four analytically different paths. A nonswitcher-nonmover remains in fossil-fuel work and remains in, or returns to, the origin commuting zone. A switcher-nonmover leaves fossil fuel but stays in the local area. A nonswitcher-mover changes local labor markets while continuing to work in fossil fuel. A switcher-mover changes both sector and place. The names are cumbersome, but the distinctions are useful because each path changes a different part of the recovery problem.
The first path preserves the most continuity and can also preserve the most exposure. A nonswitcher-nonmover does not have to pay the direct costs of relocation or learn a new sector. They may retain industry-specific knowledge and local networks. But if the local fossil-fuel labor market is shrinking, those advantages are attached to a thinner set of jobs. The worker can be reemployed and still face a lower price for the same specialized experience.
The second path changes the sector without changing the home market. This can be an efficient adjustment when a worker’s skills are portable enough to be recognized by local employers. It avoids moving costs, disruption to housing and care arrangements, and the need to rebuild a local network. It may, however, require accepting a job that uses only part of the old skill bundle or pays less than a fossil-fuel position did before the shock. The benefit is a wider set of possible employers; the cost is that the old sector-specific premium may not travel.
The third path changes place without changing sector. A nonswitcher-mover may follow fossil-fuel work to another region, including a resource area that is still expanding. This can preserve occupational identity and reduce the need for retraining. It does not guarantee a stronger bargaining position. The destination may also be dominated by a few fossil-fuel employers, so the worker has traveled without escaping the market structure that made the original recovery difficult.
The fourth path changes both place and sector. A switcher-mover may reach the broadest set of potential employers, but also faces the largest bundle of adjustments: relocation, search in an unfamiliar market, translation of prior experience, and the construction of a new job match. A broader labor market can create opportunity, but opportunity is not the same as an offer that recognizes the worker’s previous productivity. The path may be necessary for some workers and still impose a substantial short- or medium-term earnings cost.
The study’s descriptive results show why this two-margin map matters. About 20.7 percent of separated fossil-fuel workers had left their commuting zone one year after separation, rising to 38.2 percent by year seven. That is markedly higher geographic mobility than the study observes for comparable construction, manufacturing, and utility workers. The scale of movement is evidence that local adjustment is difficult. It is not evidence that most movers found a better career.
Relocation did generally make sector switching more likely. The difference between the share of switchers among movers and the share among nonmovers was about 3–4 percentage points for fossil-fuel workers, compared with 7–12 points in the comparison sectors. The direction is intuitive: moving exposes a worker to more vacancies and may reduce search or retraining barriers. The smaller effect in fossil fuel is the important part. Even after workers move, specialized skills, licensing, and the continued presence of fossil-fuel opportunities can keep them attached to the declining sector.
Subsector differences reinforce that interpretation. Workers from drilling-wells operations were 5–10 percentage points more likely to remain in fossil-fuel employment than workers from oil-and-gas extraction or coal mining, and the gap persisted among movers. Relocation changes where the worker searches; it does not automatically change what the worker can credibly offer. If the old capability is highly specific, a new commuting zone may supply another version of the old job rather than an exit from the old industry.
The map also clarifies why the four groups should not be ranked as if they were four freely chosen products. Workers do not start with the same skills, cash reserves, family obligations, health, information, or access to transportation. The study finds that movers are younger: the share under age 35 is about 12 percentage points higher than among nonmovers. They also have shorter tenure and lower pre-separation earnings. Those characteristics may make relocation more feasible in some respects, but lower prior earnings can also mean less liquidity to finance a long search, retraining, or a period of double housing costs.
Nor are the groups separated only by age. The study reports little variation across the groups in several broad demographic measures, including sex, race, and education, even while subsector and age composition differ. That combination cautions against a simple story in which one demographic category “chooses” mobility and another does not. The observed path is the result of worker capabilities and constraints interacting with the shock and the available destinations.
Why moving can help and still hurt
The apparent paradox is straightforward once the two margins are separated. Moving can widen the set of vacancies and still lead to larger earnings losses because the workers who move may be negatively selected on prior earnings and tenure, and because relocation itself disrupts a match. A worker may move not because the destination promises a gain, but because the origin no longer offers a tolerable option. In that case, moving is a response to a collapsed local opportunity set—a push—rather than evidence of a freely chosen upgrade.
The distinction between pull and push changes how we interpret the outcome. In a voluntary move, the worker may compare expected lifetime earnings, housing costs, family considerations, and career prospects before choosing a destination. After an involuntary fossil-fuel separation, the worker may instead accept the first viable destination that preserves income or occupational identity. The fact of relocation tells us that the worker crossed a geographic boundary; it does not reveal whether the move covered its costs or improved the worker’s long-run position.
The study finds that switchers generally have more stable earnings paths than nonswitchers. This pattern is consistent with a worker leaving a sector whose demand is weakening and carrying enough transferable capability into a non-fossil match. It does not mean that switching is costless. A switcher may lose an employer premium, need new credentials, or accept a lower initial wage while learning how experience is valued in the new sector. The result says that, in this setting, those costs are often smaller than the continuing losses associated with remaining in a declining sector.
Geographic movement has a different relationship to earnings. Across switching status, movers’ cumulative losses are roughly 2.2 times those of nonmovers, and the paper reports that nonswitchers experience about 44 percent larger ten-year cumulative losses than switchers. These are group comparisons, not estimates of the causal effect of personally deciding to move or switch. A mover’s larger loss may reflect the cost of relocation, the worker’s lower starting position, the conditions of the destination, or several of these at once.
A simple example shows why the combination matters. Imagine a worker whose extraction skills are recognized by employers in both the origin and destination. Moving may preserve the worker’s occupational identity, but if both places have few employers, the worker has not gained much bargaining power. Now imagine a worker with general logistics capabilities who stays in the origin zone and moves into a non-fossil employer. That worker avoids relocation costs and may have several local alternatives, even if the first new job pays less than the lost job. The first worker moves farther but may remain exposed to sector concentration; the second moves less but may improve the range of future options. These are illustrative cases, not claims about every worker in either occupation.
The migration patterns in the study fit this logic. Large rural resource regions, including parts of Wyoming, West Texas, and Appalachia, show net outflows after fossil-fuel separation. Inflows go both to other fossil-fuel-rich regions and to metropolitan commuting zones in Texas, the Mid-Atlantic, and California. Cities can offer a more diverse set of employers, but a move to a city is not the same as a move into a competitive occupation. The destination may be broader overall while the fossil-fuel niche remains concentrated, or it may contain jobs that require a different credential or local experience.
That is why a mobility statistic should be paired with an outcome and a time horizon. The one-year relocation figure tells us that many workers leave the origin market. It does not tell us whether they are employed, what they earn, whether the new job uses their previous skills, or whether they return. The seven-year classification tells us whether a long-run sector and geography transition occurred, but it does not reconstruct every period of unemployment or every unsuccessful search along the way.
The four paths also expose a practical trade-off between continuity and option value. Staying in the original sector may protect an established skill premium for as long as local demand survives, but it can leave the worker dependent on a shrinking employer pool. Switching locally may sacrifice some immediate pay while preserving housing and social ties and gaining access to more occupations. Moving within the sector may reduce local scarcity but repeat the same bargaining problem elsewhere. Moving and switching may maximize the number of potential employers while imposing the highest match-rebuilding burden. There is no universal winner because each path trades one form of risk for another.
What the evidence can support is a conditional judgment. When a worker has transferable capabilities and a local non-fossil market with credible employers, switching without moving may reduce both search and relocation costs. When the local market has few alternatives but a destination has genuinely broader opportunities, moving may be necessary. When the worker’s skills are highly specialized, moving to another resource region may preserve employment more readily than it restores career value. These are not personal prescriptions detached from circumstance; they are ways to identify which constraint a path is likely to relieve.
The classification itself is also a reminder not to blame workers for the outcome. The study follows eventual mobility status, and unobserved factors—family obligations, health, housing, networks, information, and destination selection—can affect both the decision and the earnings path. The evidence is strongest for the existence of distinct mobility patterns and their association with different outcomes. It is weaker for saying that one worker would have recovered by choosing another path, because that counterfactual is not directly observed.
The immediate job after separation can therefore be a poor guide to the eventual result. A worker may accept a temporary position, use a move to avoid prolonged nonemployment, or remain in fossil fuel while waiting for demand to return. What looks like reattachment in the first year may be followed by a deeper earnings or employment effect as savings, severance, and temporary matches run out. To understand that delayed damage, the next step is to follow these paths through time rather than treating the first observed destination as the end of the story.