Generally, the numerous numbers included within the authorities’s month-to-month jobs report come collectively to color a transparent, coherent image of the power or weak point of the U.S. labor market.

This isn’t a kind of occasions.

As an alternative, the information launched by the Labor Division on Friday was a large number of conflicting alerts. It couldn’t even agree on essentially the most fundamental of questions: whether or not the financial system is including or shedding jobs.

The report confirmed that employers added 272,000 nonagricultural jobs in Might, excess of forecasters have been anticipating. That determine is predicated on a survey of about 119,000 companies, nonprofit organizations and authorities companies.

However the report additionally accommodates knowledge from one other survey, of about 60,000 households. That knowledge confirmed that the quantity of people that have been employed final month really fell by 408,000, whereas the unemployment charge rose to 4 % for the primary time in additional than two years.

The 2 surveys measure barely various things. The employer survey consists of solely workers, for instance, whereas the family survey consists of unbiased contractors and self-employed staff. However that doesn’t clarify the discrepancy final month: Adjusting the family survey to align with the ideas used within the employer survey makes the job losses in Might look bigger, not smaller.

That signifies that the conflicting footage come right down to some mixture of measurement error and random noise. That’s irritating however common: Over the long run, the 2 surveys typically inform related tales, however over shorter intervals they continuously diverge.

Economists sometimes put extra weight on the employer survey, which is way bigger and is usually considered as extra dependable. However they aren’t positive which knowledge to consider this time round. Some economists have argued that the family survey could possibly be failing to seize totally the current wave of immigration, main it to undercount employment progress. However others have argued that the employer survey could possibly be overstating hiring as a result of it isn’t accounting correctly for current enterprise failures, amongst different elements.