On-the-Job Search

Workers who remain employed while looking for their next role—“on-the-job search”—have become a dominant force in modern labour markets. This mechanism drives job mobility, wage growth, and the concentration of wage increases among those already in work, reshaping the path of wage inequality and economic efficiency.

Most job-to-job transitions start while employed

In most modern economies, the majority of workers who change jobs do so while already holding a position. They send out applications, interview on Tuesdays, and negotiate offers without ever appearing on an unemployment roster. This shift has profound consequences. An employed job-seeker has bargaining power: employers know the recruit has an outside option, and the worker has maintained earnings, references, and workplace credentials. By contrast, unemployed searchers face a handicap—employer suspicion, skill degradation during joblessness, and financial pressure to accept any offer quickly.

The prevalence of on-the-job search varies by education, age, and labour market tightness. Highly skilled professionals in tight labour markets (technology, finance) may secure new roles before resignations; lower-wage workers in slack labour markets must often remain unemployed between jobs. Yet across OECD economies, on-the-job transitions now account for 50–70% of all job changes, a rise from lower shares in the 1990s. This shift reflects both worker preference—staying employed is safer—and employer behaviour: firms now hire laterally from competitors rather than training unemployment recruits.

On-the-job search concentrates wage growth among the employed

Because those already working can search selectively, they tend to negotiate higher wage increases than those hired from unemployment. An on-the-job searcher can wait for a match with better pay, benefits, or career trajectory; an unemployed applicant is often desperate. Studies of job-to-job transitions show wage gains of 5–15% upon switching, with larger gains for workers moving to expanding firms or occupations. This same mechanism explains why wage growth correlates more closely with job-switching rates than with unemployment: when workers move freely between employers, wages rise; when on-the-job search is chilled by recession or labour market tightness, aggregate wage growth slows.

The distributional consequence is significant. In tight labour markets where on-the-job search thrives, employed workers capture most gains while unemployment rates fall but labour-force participation may stagnate. The unemployed and underemployed benefit less from the wage pressure. This dynamic partly explains rising wage inequality in recent decades: workers who already have jobs enjoy accelerating wage growth from job mobility, while those starting from unemployment face deteriorating relative prospects.

Poaching and wage-offer generation

When firms find it easier to hire experienced workers from competitors than to train new ones, on-the-job search becomes the target. Tech companies, for instance, spend heavily on recruitment teams trained to identify and approach employed engineers, offering signing bonuses and stock grants. This behaviour—called “poaching”—raises wage growth in the poached industry but can reduce capital investment if firms spend recruitment budgets instead of training.

Wages also rise when on-the-job search is active because firms must refresh offers to retain staff. An employer aware that a talented employee is receiving outside offers may counter-offer to keep them. In models of wage setting, this threat of worker exit drives wages closer to each worker’s marginal revenue product, rather than a mere reservation wage. The presence of on-the-job search thus tightens the link between productivity and pay.

Trade-off: efficiency gains and coordination costs

On-the-job search accelerates worker reallocation toward more productive employers and roles, improving aggregate productivity. A researcher hired by a rival lab may do better work there; a sales representative with a rival firm may find a better product-customer fit. Labour economists generally view this reallocation as efficiency-enhancing.

But there are coordination costs. If searching is rampant, firms may spend heavily on retention bonuses and recruitment to prevent talent drain, deadweight expenses that don’t raise output. Furthermore, when many workers are searching, information asymmetries abound: employers cannot easily distinguish job-hoppers from committed employees, or assess true candidate quality before hiring. This can lead to wage compression (employers paying all new hires a safe middle rate) or quality screening (firms trusting only employee referrals, which can exclude outsiders).

Cyclical fluctuations and recession

On-the-job search is procyclical: it accelerates in expansions and shrinks in downturns. In a boom, employed workers sense wage opportunities and bid up offers; in recession, fear of job loss and scarce vacancies chill search intensity. The Great Recession saw a sharp drop in job-to-job flows, with unemployed workers bearing the adjustment. Recovery was slow: even as unemployment fell, on-the-job search and wage growth remained subdued for years, a phenomenon sometimes called “wage scarring” from the crisis.

This cyclicality matters for monetary policy. Central banks watch job-to-job transition rates as a leading indicator of inflation pressure. A pickup in on-the-job switching suggests worker confidence and wage demand; persistent weak switching during growth may signal slack labour supply and permit easier policy.

On-the-job search and the wage curve

The wage curve — the empirical pattern that wages are lower in high-unemployment regions — partly reflects differences in on-the-job search intensity. In regions with slack labour markets, fewer workers can afford to search while employed, and more must transition through unemployment. This shifts the wage-setting process, pinning wages closer to unemployment insurance or reservation wages. In tight regions, on-the-job search is more abundant, wages drift higher. This mechanism underpins why local unemployment is so strongly correlated with local wage levels.

See also