Ending pandemic unemployment benefits early was supposed to make people go back to work faster, said Kristi Noem and other mostly Republican pandemic deniers. Early hiring data indicated Noem et al’s scapegoating was economically incorrect. So does state payroll data for July. States that cut unemployment benefits in June saw little change in job growth from June to July, while states still issuing enhanced UI saw a bigger jump in job growth and stronger growth than the stingy states:
…The simplest view can be seen when comparing job growth in states that ended programs early against those that did not. The former group experienced aggregate gains that were about 10 basis points lower than those that made no changes. Although this gap is not large enough to reflect a meaningful difference, the fact that growth was weaker in states that cut benefits makes it clear that the hoped-for flood of worker re-entries into the labor force has not materialized.
…Though this could have reflected an anticipation of a policy change, there is little anecdotal evidence that behavior meaningfully changed following initial announcements. Further, the acceleration in job gains across the nation from June to July suggests that there was no onetime spark around the time of implementation. In fact, the first four states to cut benefits—Alaska, Iowa, Mississippi, and Missouri—were generally middling to below-average performers in June.
All told, these results suggest a negative relationship between cutting UI benefits and job growth [Adam Kamins, “Covid-19 Surges Matter, UI Cuts Don’t, So Far,” Moody’s Analytics: Weekly Market Outlook, 2021.08.26, p. 4].
The four states with the fastest job growth in July and seven of the top ten kept the extra UI payments in place. This economic experiment ends this week, when the extra unemployment benefits approved by Congress to get people through the pandemic end on September 4. But we are emerging from this experiment with a couple months’ worth of data showing that when Kristi Noem makes economic recommendations, we should do the opposite.