This morning’s jobs report, showing an April surprise of 288,000 jobs added, was welcome news. This is the kind of job growth that would get us back to health in the labor market relatively soon. If we kept up this pace, we would get back to pre-recession labor market conditions by the end of 2016.

However, this was a particularly wild case of “a tale of two surveys.” The household survey showed a substantial drop in the unemployment rate—to 6.3 percent—but the drop was entirely due to people dropping out of the labor force. Employment in the household survey actually declined, and the labor force participation rate fell back down to its lowest point of the recovery. Our estimate of the number of “missing workers” (workers who are not working or actively seeking work but who would be if job opportunities were strong) increased to an all-time high of 6.2 million. If those missing workers were in the labor force looking for work, the unemployment rate would be 9.9 percent instead of 6.3 percent.

As always, when the two surveys tell different stories, the rule of thumb is to place much more weight on the payroll survey, since it is much larger and less erratic. However, the weak household survey hugely dampens the enthusiasm of the strong payroll numbers. We can’t be confident that the economy is entering a new stage of stronger job growth until both surveys are regularly posting strong gains, and that did not happen in April.