💼 Labor Market Job Search (McCall)

McCall job search: a worker receives wage offers w ~ F(w) and accepts if w ≥ w* (reservation wage). Bellman equation w* = b + β∫max(w,w*)dF determines optimal stopping.

EconomicsInteractive
Simulate Search to run one worker · Adjust sliders to shift reservation wage

How it Works

In the McCall (1970) job search model, an unemployed worker receives one wage offer per period drawn from distribution F(w). The worker decides each period whether to accept (earning that wage forever) or reject (receive benefit b, draw again next period).

The optimal policy is a reservation wage w*: accept if w ≥ w*, reject otherwise. The Bellman equation defines w* as the wage that makes the worker indifferent between accepting and continuing search:

Value of accepting w: V(w) = w / (1 - β) Value of continuing: U = b + β · E[max(V(w), U)] Reservation wage w*: w* = (1-β)·U = b·(1-β) + β·∫_{w*}^∞ (w - w*)·dF(w) Expected duration: E[T] = 1 / (1 - F(w*))

The simulation plots the wage offer distribution with the reservation wage threshold. Each drawn offer is shown as a dot; rejected offers appear in red, the accepted offer in green. The histogram of offers accumulates as search progresses.

Frequently Asked Questions

What is the McCall job search model?

The McCall (1970) job search model describes an unemployed worker who receives one wage offer per period drawn from a known distribution F(w). The worker decides to accept (ending search) or reject and continue searching. The optimal strategy is a reservation wage rule: accept if w ≥ w*.

What is the reservation wage?

The reservation wage w* is the minimum wage a worker will accept. It is determined by the Bellman equation: the value of accepting w* equals the value of continuing to search. All offers above w* are accepted; all below are rejected.

What is the Bellman equation in job search?

The Bellman equation states: w*/(1-β) = b/(1-β) + β∫max(w, w*)dF(w), where b is unemployment benefit, β is the discount factor, and F is the wage offer distribution. This defines the optimal stopping point.

How does the discount factor affect search?

A higher discount factor β (closer to 1) means the worker is more patient and places more value on future wages. This raises the reservation wage and prolongs search. A very impatient worker (low β) accepts almost any offer.

How do unemployment benefits affect the reservation wage?

Higher unemployment benefits b increase the opportunity cost of accepting a low wage, raising the reservation wage. This extends unemployment duration but leads to better job matches — the classic efficiency-insurance trade-off.

What determines expected unemployment duration?

Expected duration = 1/(1-F(w*)). If the reservation wage is high, only rare good offers are accepted, and average duration grows. The model predicts that duration increases with benefits and with the dispersion of the wage distribution.

What is optimal stopping theory?

Optimal stopping is the mathematical problem of choosing the best time to take an action (accept an offer) given a sequence of random draws. The McCall model is a classic application: the optimal rule is a threshold (reservation wage) above which you stop.

What wage distribution does the model use?

The original McCall model uses any distribution F(w). Common choices include uniform, log-normal, and normal distributions. Log-normal captures the right-skew of real wage distributions. The reservation wage shifts with the mean and variance of F.

What is job matching theory?

Job matching theory (Mortensen, Pissarides) extends McCall by adding firms searching for workers. The equilibrium wage is determined by Nash bargaining over the joint surplus. This creates the matching function linking vacancies, unemployment, and hires.

Can the reservation wage change over time?

In the basic McCall model with infinite horizon, w* is constant. With finite unemployment benefit duration (benefit exhaustion), the reservation wage falls as the deadline approaches — workers become less choosy, consistent with empirical evidence on job finding spikes near benefit exhaustion.

About this simulation

This is McCall's 1970 optimal-stopping model of unemployment: a worker draws one random wage offer per period from a chosen distribution and must decide, purely from the numbers, whether to accept it forever or reject and try again next period. The simulation solves the Bellman fixed point for the reservation wage w* and then plays out a live search, colouring each rejected offer red and the eventual accepted offer green.

🔬 What it shows

The wage-offer probability density curve with the accept region shaded to the right of a dashed reservation-wage line w*, plus a running column of sampled offers stacked as dots — red for rejected, green for the one finally accepted.

🎮 How to use

Adjust unemployment benefit b, discount factor β, wage mean μ, and wage spread σ with the sliders, switch between Normal, Log-Normal, and Uniform wage distributions, then click Simulate Search to watch one simulated worker search period by period until an offer clears w*.

💡 Did you know?

Raising the discount factor β toward 1 makes a worker act more patient and pushes the reservation wage up — but it also mechanically lengthens expected search duration via E[T] = 1/(1−F(w*)), which is exactly why generous, long-lasting unemployment benefits are predicted to extend unemployment spells even for perfectly rational searchers.

Frequently asked questions

Why doesn't the worker just accept the first offer above the mean wage?

The reservation wage w* is set by the Bellman equation to make the worker exactly indifferent between accepting now and continuing to search, accounting for the discounted expected value of future draws — so w* can sit meaningfully above or below the mean depending on b, β, and the spread of the distribution.

Why does raising unemployment benefit b push the shaded accept region rightward?

Benefit b is the guaranteed payoff from rejecting an offer, so a higher b raises the value of continuing to search, which raises w* in the Bellman fixed point — visually the dashed reservation-wage line and its shaded region shift right as you increase the slider.

What does switching to Log-Normal change about the search?

Log-normal wages are right-skewed, so for the same mean μ there's a longer tail of rare high offers and more mass at low wages compared to Normal — this changes both the shape of the plotted PDF and the computed w*, since the model integrates the tail above w* to find the Bellman fixed point.

Why does search sometimes take many periods and sometimes very few?

Because offers are drawn randomly each period, the actual duration in one run is a geometric random variable with success probability 1−F(w*) — the "E[duration]" stat shows the theoretical average, but any single Simulate Search run can finish quickly or drag on by chance.

Is a higher reservation wage always better for the worker?

Not unconditionally — it trades off shorter search against a better eventual wage match, the classic efficiency-insurance trade-off: a stingier benefit b pushes w* down and shortens unemployment, while generous benefits raise w* and stretch out the expected duration shown in the stats panel.