What a Real Recovery Requires

For displaced workers, real recovery begins where the reemployment statistic ends: when a new job becomes durable work, restores a credible share of prior earnings, uses enough of the worker’s capabilities to preserve future options, and leaves the worker with alternatives when the next shock arrives. The final judgment of this series is therefore practical as well as empirical: measure more than a job return, diagnose which constraint is binding, and match assistance to that constraint without treating the fossil-fuel case as a forecast for everyone.

The distinction is easy to state but demanding to apply. The Bureau of Labor Statistics reports that 66.1 percent of long-tenured workers displaced between 2023 and 2025 were reemployed in January 2026. That is an important national measure of reattachment. It says much less about how long workers were without earnings, whether the new job paid as much, whether their skills still command a premium, or whether the local market gives them a credible alternative to the current employer. The fossil-fuel evidence makes those missing questions measurable: workers move unusually often, earnings losses accumulate over years, and destination concentration determines whether a skill switch becomes a bargaining gain.

A recovery test, not a job count

A useful recovery test starts with employment attachment. A person who has no job at the survey date is not in the same position as someone who has returned to work. Yet “reemployed” is also not a sufficient endpoint. The BLS release’s own earnings comparison shows why: among a defined subgroup of long-tenured workers who lost full-time wage-and-salary jobs and were reemployed in January 2026, about 49 percent earned as much or more than on the lost job. The comparison is not for every displaced worker, and it does not measure every feature of job quality, but it establishes the essential distinction. Job return and earnings restoration are separate outcomes.

The next test is persistence. A worker who records one new job may still be in a fragile transition. The job can be temporary, seasonal, or a poor match that ends quickly. Even a stable job can leave the worker with fewer hours, weaker benefits, or less access to the next step on the job ladder. The BLS snapshot cannot reconstruct those sequences. The historical fossil-fuel study shows why the sequence matters: local nonemployment effects are small before exposure, rise after the shock, and grow to more than one percentage point by about year ten in the local-market estimates. A recovery measure that observes only the first landing can miss damage that appears after temporary supports and initial matches have run out.

A third test concerns earnings over time rather than earnings at one moment. The fossil-fuel study reports an initial period in which earnings effects are mixed or modest, followed by a more consistent decline around years three through six and stabilization at a lower level in years seven through ten. It summarizes the average relationship as roughly a 0.16 percent annual earnings decline for a one-percent decline in national fossil-fuel employment, with an interquartile exposure difference corresponding to about a 2.5 percent annual earnings loss. At the peak, that interquartile difference corresponds to an average annual increase of approximately ten weeks in nonemployment. The magnitudes belong to the study’s exposure scale and historical sample; the general lesson is the timing. A career can be damaged after a worker has already returned to work.

A fourth test asks what happened to the worker’s capabilities. The new job may use the old skill bundle, translate only part of it, or disregard it altogether. That distinction affects both current pay and future option value. A worker who leaves fossil fuel for an adjacent occupation may accept a temporary wage cut while retaining technical knowledge that becomes valuable later. A worker who accepts a job that uses none of the old capability may regain current income but lose the opportunity to build on prior experience. Administrative employment records can reveal earnings and sector transitions, but the detailed study does not directly observe every worker’s occupation, so its subsector comparisons are proxies for portability rather than a complete skill inventory.

A fifth test is geographic and institutional.