Measuring who moves
Identifying which workers change employers — and how often — requires data that follows individuals across jobs over time. Two of the most widely used sources for this purpose are the Current Population Survey's tenure supplement, which asks respondents how long they have been with their current employer, and the Longitudinal Employer-Household Dynamics (LEHD) program maintained by the Census Bureau, which links state unemployment insurance records to construct quarterly employment histories for the majority of the U.S. workforce.
Melissa Bjelland, Bruce Fallick, John Haltiwanger, and Erika McEntarfer used LEHD data in a 2011 paper published in the Journal of Business & Economic Statistics to produce comprehensive estimates of employer-to-employer flows in the United States. Their analysis covered the period from 1998 to 2003 and provided one of the first large-scale portraits of who moves directly from one employer to another. The study found that direct job-to-job transitions are concentrated among younger workers, workers in industries with high baseline turnover such as retail and food services, and workers in the early years of a given employment spell.
Age, education, and industry
The relationship between age and job mobility is among the most robust findings in labor economics. Topel and Ward's 1992 analysis of young men's careers documented that the rate of employer changes declines sharply with age, particularly after the first several years in the labor force. This pattern has been replicated in numerous subsequent studies and across different data sources. Workers in their twenties change employers at roughly twice the rate of workers in their forties, though the precise magnitudes vary by dataset and definition of a job change.
Education interacts with job mobility in ways that are less straightforward than the age gradient. Workers with college degrees tend to have lower raw switching rates than workers without degrees, partly because college-educated workers are more likely to be in salaried positions with longer average tenure. However, when college-educated workers do switch, they are more likely to do so through a direct job-to-job transition rather than through an intervening spell of unemployment, and the wage gains associated with their switches tend to be larger in percentage terms.
Industry composition plays a significant role as well. Sectors characterized by project-based work, seasonal demand, or high rates of entry and exit — including construction, hospitality, and parts of professional services — exhibit switching rates well above the national average. Government employment, by contrast, is associated with substantially lower turnover at all age levels.
A long-run decline in dynamism
Henry Hyatt and James Spletzer documented in a 2013 IZA Journal of Labor Economics paper that employer-to-employer transition rates in the United States fell markedly between the late 1990s and the early 2010s. The decline was broad-based, appearing across age groups, industries, and geographic regions, though it was most pronounced among younger and less-educated workers. Raven Molloy, Christopher Smith, and Abigail Wozniak reported a parallel trend in their 2011 Journal of Economic Perspectives article on internal migration: Americans were moving across state lines less frequently than in previous decades, and the decline in geographic mobility appeared closely linked to the decline in job mobility.
The causes of this trend remain actively debated. Proposed explanations include an aging workforce, increased occupational licensing requirements that raise the cost of switching, greater employer consolidation reducing the number of viable alternative employers in local labor markets, and improvements in employer-employee matching that reduce the number of poor matches that might otherwise lead to early exits. No single explanation has achieved consensus, and the relative contribution of each factor likely varies across segments of the labor market.
Limitations of the evidence
Longitudinal datasets are powerful but imperfect instruments for studying job mobility. Panel surveys such as the National Longitudinal Survey of Youth (NLSY) can track the same individuals for decades, but they are subject to attrition — respondents who leave the sample may differ systematically from those who remain. Administrative datasets like the LEHD avoid survey attrition but lack information on hours worked, occupation, and education for many records, making it difficult to fully characterize the workers who switch.
Self-reported tenure data in the CPS carry known measurement error. Respondents tend to round their reported tenure to whole years, and the question wording has changed over time, introducing breaks in the series. Researchers who rely on CPS tenure data for mobility estimates typically adjust for these issues, but the adjustments themselves introduce assumptions that may not hold uniformly across demographic groups.