“… the report shows an increase in the employment-population ratio, the percentage of adults who are working. What?
OK, the trick is that there are two different surveys. Payroll numbers come from a survey of firms; unemployment (and employment-population) numbers come from a survey of households. Both surveys are subject to error, both strict statistical sampling error and things like incomplete coverage, uncertain seasonal adjustments, and so on. When employment growth is near zero, on either side, it’s not that surprising that the surveys should point in opposite directions.”
