This article from Bloomberg about the variability of hours worked by those who have had to accept lower paying jobs is instructive not only for the skepticism we should give to the assumed benefit rising employment statistics, but also to the human factor in what affects "productivity." In the cold view of traditional economics if productivity doesn't rise with real wage rate increases then that is bad (feather bedding employees!). Something in the system has to be punished; whether by rising interest rates, or pressure to lower wages, or both. And it may well be that it's not just workers who get the brunt end of the stick, but you can bet that the major holders of capital don't mind so much. They like to charge more for lending their counters, or to have better returns on big deposits.
Just more to keep in mind when you hear about singular economic statistics.