I have been monitoring the information trickling out about the effect of dramatically slashing unemployment insurance (UI) benefits in twenty-six states. My first post used the Census Pulse Survey to show that employment did not grow faster (indeed it grew slower) in the twelve states that cut UI benefits in early June. My second post reported on a study using bank transaction data that showed that individuals who were on UI in April became a lot poorer in states where UI was cut but only very negligibly increased their employment.

In this post, I look at the BLS state payrolls data, which was released late last week. For this analysis, I break states into four categories: the twelve states that cut UI between June 12 and June 19 (“Early June”), the thirteen states that cut UI between June 26 and July 3 (“Late June/Early July”), the one state that cut UI on July 31 (“Late July”), and the remaining twenty-four states that are scheduled to cut UI on September 6 (“Early Sept”).

From there, I looked to see how payroll levels changed in each group relative to May of this year, the month before UI cuts began. If these cuts were juicing employment, you would expect to see payrolls increasing the most in the Early June states and the least in the Early Sept states, with the Late June/Early July and Late July states somewhere in between.