Every labour statistic in the world is built from a partition of the working-age population into three boxes, and almost every mistake comes from forgetting which box somebody is in.
Take the population above a stated age, usually 15 or 16, excluding people in institutions and, in most definitions, the armed forces. Every person in it is exactly one of three things. Employed: did any work for pay or profit in the reference period, including part-time work and unpaid work in a family enterprise above a threshold of hours. Unemployed: not employed, available for work, and actively searched during a stated recent window, usually four weeks. Not in the labour force: everybody else, including students, homemakers, the retired, the long-term ill, and people who want a job and have stopped looking.
How is the labour force participation rate calculated?
The labour force is the employed plus the unemployed. Four ratios follow, and each answers a different question:
| measure | formula | what it answers |
|---|---|---|
| Labour force participation rate | labour force ÷ working-age population | how much of the population is engaged with the labour market at all |
| Unemployment rate | unemployed ÷ labour force | how much of the engaged population cannot find work |
| Employment-population ratio | employed ÷ working-age population | how much of the population is actually working |
| Underemployment or broad rate | (unemployed + marginally attached + involuntary part-time) ÷ (labour force + marginally attached) | how much labour is idle or under-used, not just openly jobless |
The three rate measures are connected by an identity worth carrying: the employment-population ratio equals the participation rate multiplied by one minus the unemployment rate. Use it as a check on any panel you are handed, because a panel that fails it has a definitional problem somewhere.
Advertisement
What are discouraged workers and underemployment?
Two categories inside the third box do most of the damage to naive readings. A marginally attached worker wants a job and is available for one but did not search in the reference window, so is counted as outside the labour force. A discouraged worker is the subset of those whose stated reason for not searching is a belief that no work is available.
Neither appears in the unemployment rate. Both appear in a broad rate, and the gap between the two is a better cyclical indicator than either alone.
Underemployment is a third idea and a different one. It covers people who are employed but under-used: working part-time because full-time work was not available, or working far below their skill level. It matters enormously in economies with thin safety nets, where almost nobody can afford to be openly unemployed, so distress shows up as low-productivity self-employment rather than as a rising unemployment rate.
A headline unemployment rate of 3.32% on a participation rate of 43.46%, with 13.96% of the employed reporting they want more work, describes an economy with a serious labour problem and a flattering headline.
Why can the unemployment rate mislead?
The practical warning follows directly from the boxes. The unemployment rate has a denominator that moves, and it moves for good reasons and bad ones. People entering the labour force raise it. People leaving lower it, by shrinking the denominator rather than by finding work. Worked Example 3 computes both directions on real-shaped panels, and the swing is large enough to reverse the sign of the story.