Permabulls Brian Wesbury and Robert Stein argue that because the household survey say a much bigger gain (422,000) in jobs than the payroll survey's (69,000), the payroll survey underestimates the strength of the U.S. job market. That might have been plausible if that had been a consistent pattern, but that is not the case as the household survey showed significant drops in employment in March and April even as the payroll survey showed gains. In fact for the last 3 months as a whole, the payroll survey showed a bigger increase (279,000) than the household survey (222,000).
This illustrates that the household survey is more erratic and volatile than the payroll survey and that with regard to short-term economic trends, the payroll survey is more reliable. The household survey number could however be useful if the household survey is consistently either stronger or weaker than the payroll survey than that would be reason to suspect that the payroll survey either underestimates or overestimates strength in the job market. However, when it comes to short-term economic trends, which is what Wesbury's and Stein's comments were about, then the payroll survey is more reliable.