[Dave W]

The Census Bureau on behalf of the Labor Dept calls about 65,000 households a month. They ask everyone in the household if they have a job. If they don't have a job, they ask if they have looked for work in the past 4 weeks. If they have not, then they are discouraged job seekers and are taken out of the labor market. I'm not exactly certain what constitutes "looking for work", but certainly sending resumes, etc. should qualify as looking for work. This is the survey that actually calculates the unemployment percent.

There is another survey called the business survey and this one determines the number of jobs added or lost during the month.

So it is possible that the household survey can suggest that jobs were added in the past month because it captures people starting businesses from home, etc, while the business survey indicates many less jobs added since this survey does not capture any new businesses started. The surveys quite often conflict with each other, but over a longer period of time, they come together.

The household survey quite often captures a recovery faster than the business survey. The business survey will catch up over time

[existentialist]

Thanks. WOW, what an odd way of coming up with the unemployed percentage, but I suppose if it’s done the same each month a trend can be established.