Labor Hoarding

Labour hoarding is the deliberate retention of more workers than a firm needs during a recession or downturn, betting that business will rebound before long. By keeping workers on payroll (sometimes at reduced hours or wages), firms avoid rehiring costs, skill loss, and the risk of losing workers permanently to competitors. The result is a disconnect between falling output and stable employment—unemployment rises less sharply than economic contraction would suggest.

Why firms choose to hoard rather than cut

Firing and rehiring is expensive. When a worker leaves—whether by layoff or resignation—the firm must post a job, recruit, interview, train, and wait for a replacement to reach full productivity, which can take weeks or months. New hires often make costly mistakes. If a downturn is forecast to last only two to four quarters, hoarding workers costs less than the rehiring treadmill. A manufacturing firm cutting 15% of its workforce in a mild recession may struggle to scale back up if demand snaps back; that firm might instead negotiate shorter hours, temporary pay cuts, or unpaid leave, keeping bodies on the payroll and skills in house.

Beyond costs, hoarding reflects confidence. If management believes the downturn is cyclical and brief, holding workers signals to them—and to the market—that the firm expects to rebound. Workers are less likely to quit a hoarding firm, sensing stability ahead. Competitors face the same calculation: if everyone hoards, everyone’s labour costs stay high, and no one gains a wage-cutting advantage. In tight labour markets before the downturn, firms have already invested in recruiting and training. Throwing that away feels wasteful.

Hoarding is strongest in sectors where workers are difficult to replace: skilled trades, engineering, healthcare. It is weaker in low-skill, high-turnover sectors like food service and retail, where replacing workers is routine and quick.

The productivity paradox

Labour hoarding creates a strange corporate snapshot: output is falling, but employment barely moves. This means productivity per worker collapses. In a typical recession, GDP falls 2–4% while employment drops 1–2%; in a hoarding-heavy downturn, employment might fall only 0.5%. The imbalance is made up in productivity: workers produce less per hour, work shorter weeks, or sit idle waiting for demand to return. From a national-accounts perspective, this shows up as a sharp drop in labour productivity, which can confuse observers. The economy is not becoming less efficient; firms are simply choosing to carry excess labour capacity as an insurance policy.

This trade-off becomes unsustainable quickly. If a downturn lasts longer than expected, hoarding firms either capitulate and lay off en masse, or they attempt to reduce wages and hours further. Japan in the 1990s took the extreme route, keeping workers employed but on reduced hours and frozen wages for years—a choice that dampened unemployment but also inflation, wage growth, and consumer demand, extending the slowdown.

The effect on unemployment and employment-cost-index

Labour hoarding is why unemployment often lags GDP declines. In the 2008–2009 financial crisis, U.S. GDP fell 4.3% at its trough, but unemployment peaked at 10%. If firms had adjusted employment proportionally to output, unemployment might have hit 12–13%. Instead, hoarding kept joblessness lower in the first two quarters of the downturn, then employment collapsed as firms realized recovery would be slow. By contrast, in faster downturns like the early 1980s recession, employment fell more sharply upfront because firms cut quickly and rehired just as fast.

The Employment Cost Index is dampened by hoarding too. Firms holding onto workers are less likely to raise wages, knowing labour supply is available. Hours may fall, and compensation per worker may stagnate. The ECI grows more slowly in hoarding economies than in those with rapid labour adjustment. This can mislead policymakers: a flat ECI during a downturn might signal lack of wage pressure, when in fact firms are simply not bidding for workers because they already have excess supply.

Hysteresis and the long-term cost

Labour hoarding is a short-term buffer, not a long-term strategy. If a recession becomes a depression—or if a cycle of recessions repeats—workers lose skills, confidence, and connections. Those retained in hoarding jobs may not be the right workers for post-recession growth; firms struggle to redeploy them, and workers’ bargaining position weakens. This dynamic feeds hysteresis in unemployment: a temporary downturn where firms hoarded becomes a permanent scar where once-employed workers drift into long-term joblessness, skill decay, and discouraged worker effects.

The COVID-19 downturn saw aggressive hoarding in many sectors, especially hospitality and aviation, where firms knew lockdowns were temporary and labour shortages were imminent post-recovery. Those firms that hoarded aggressively were rewarded; those that cut were forced to poach workers at premium wages. But hoarding only works if recovery is genuine and soon. Prolonged downturns—or waves of repeated shocks—exhaust its value.

When hoarding breaks down

Firms stop hoarding when cash runs out, when losses mount beyond tolerance, or when recovery expectations shift from “soon” to “never.” In the 2008 crisis, hoarding held through Q1 2009; by Q2 and Q3, massive layoffs began as firms accepted that the recovery would be long. In European recessions, stronger labour-protection laws and union agreements extend hoarding further, sometimes for years, keeping headline unemployment artificially low while underemployment and discouraged workers rise.

The decision to stop hoarding is often sudden and brutal. A firm holding workers through quarter three of a downturn abruptly announces severance and closure. The labour market absorbs a shock of supply far exceeding what gradual layoffs would have produced. This creates tail risk for policymakers: hoarding obscures true labour-market slack during downturns, then releases it suddenly, causing sharp unemployment spikes and policy errors.

See also