4.5.2—Measuring unemployment

The claimant count records people claiming an unemployment-related benefit, while a labour force survey asks a sample about work, availability and job search. Neither measure is a perfect census of every person without work.

Claimant data are relatively frequent and administrative, but eligibility rules and benefit incentives can change the series. Survey data apply a consistent labour definition more directly, yet sampling error, non-response and recall problems affect estimates.

If benefit rules become stricter, claimant unemployment may fall even when survey unemployment does not. A worker who wants a job but has stopped searching may be missed by an active-search definition and classified as inactive.

A change in the measured rate need not equal a change in the underlying labour market. Check the definition, coverage and method before comparing periods or countries.