Unemployment rate

A statistic compiled by a national statistics agency from a survey of households. A person is counted as unemployed only when three conditions hold together: without work, available to start, and having actively searched within a defined reference period. The labour force is the employed plus the unemployed, and the rate is the second divided by the first. Anyone not searching is outside the labour force entirely and appears in neither part of the fraction.

That last exclusion is the whole of the arithmetic's subtlety. Because people leaving the labour force shrink the denominator without adding to the numerator, the rate can fall in a month when nobody gained a job, and it can rise in a month when employment grew, if discouraged workers resumed searching faster than jobs were created. This is why statistics agencies publish the participation rate beside the headline, along with broader measures that include people working part time who want full time work, and why the figures are seasonally adjusted and later revised.

Markets read it through the central bank, since a labour market's tightness shapes what the policy rate is expected to do, and the reaction comes from the gap between the reading and the published consensus rather than from the level. Two traps follow. In the United States the rate is released in the same report as non-farm payrolls, but the two come from different surveys, one of households and one of employers, so they can point in opposite directions in the same release without either being wrong. And the relationship between unemployment and inflation, long taught as a stable trade off, has behaved inconsistently enough over recent decades that economists disagree openly about how much a given reading implies for rates at all.