Labor Markets: Search Frictions, Tightness and the Beveridge Curve

A market that never fully clears

Unlike a textbook market where price adjusts instantly to match supply and demand, the labor market always has both unemployed workers and unfilled vacancies at the same time — search and matching frictions mean a willing worker and a hiring firm cannot find each other instantly. The Diamond-Mortensen-Pissarides (DMP) model, whose developers shared the 2010 Nobel Memorial Prize in Economics, formalizes this friction with a matching function that turns the stock of unemployed workers u and vacancies v into a flow of new hires:

m = m(u, v)          new matches per period, increasing in both u and v

The ratio θ = v/u is labor market tightness — how many vacancies exist per unemployed worker. When θ is high, firms compete over a relatively scarce pool of job-seekers and vacancies take longer to fill; when θ is low, workers compete over relatively scarce openings and unemployment spells lengthen.

Job finding and job filling rates

From the matching function, two rates fall out directly: the job finding rate f(θ) = m/u, the probability an unemployed worker finds a job in a given period, and the job filling rate q(θ) = m/v, the probability a vacancy gets filled. Both depend on tightness θ, and in opposite directions — a tighter market (high θ) raises the job finding rate for workers but lowers the job filling rate for firms, exactly the congestion effect you'd expect: more competition for whichever side is scarce.

Inflow meets outflow: the steady state

Unemployment does not just depend on hiring; it also depends on separations — workers losing or leaving jobs at some rate s per period. In steady state the flow of workers into unemployment (separations from the pool of employed workers) exactly balances the flow out of it (successful job finding), which pins down an equilibrium unemployment rate as a function of the separation rate and the job finding rate. A labor market can reach a stable unemployment rate even while individual workers are constantly cycling in and out of jobs — steady-state unemployment is a balance of flows, not a fixed group of permanently unemployed people.

The Beveridge curve

Plotting the unemployment rate against the vacancy rate over time traces the Beveridge curve, which normally slopes downward: when the economy is booming, vacancies are plentiful and unemployment is low; in a downturn, vacancies dry up and unemployment rises. Matching efficiency shows up directly on this plot — an outward shift of the whole curve (more unemployment and more vacancies existing simultaneously) signals that the matching process itself has gotten less efficient, workers and openings are having a harder time finding each other for reasons other than the aggregate level of economic activity, such as skill mismatch or geographic mismatch between where workers live and where jobs are.

Why this model shapes real policy

Because the DMP model explicitly separates how many jobs exist from how efficiently workers and jobs find each other, it gives policymakers a way to diagnose unemployment: is it because aggregate demand is weak and there simply aren't enough vacancies (a job-creation problem), or because matching efficiency has fallen and job-seekers and openings are failing to connect despite jobs existing (a search-friction problem)? These call for very different policy responses — stimulus versus retraining, relocation assistance or better job-matching infrastructure — and the Beveridge curve's shape is one of the main diagnostic tools economists use to tell the two apart in real time.

Frequently asked questions

Try it live

Everything above runs in your browser — open Labor Market and change the parameters while it is running. Nothing is installed, nothing is uploaded, the whole model lives in one tab.