Steven Mnuchin warned senators in mid-March 2020 that U.S. unemployment could reach about 20% without swift economic rescue measures in response to COVID-19 shutdowns, a scenario widely reported as a worst‑case outcome. (cnbc.com)

In reality, the official U.S. unemployment rate (U‑3) peaked at 14.7% in April 2020 and then declined to 13.3% in May, 11.1% in June, 10.2% in July, and 8.4% in August—well below the 20–30% range “by the summer of 2020.” (jobmarketmonitor.com) A broader “real” unemployment measure (U‑6) did briefly reach about 22.8% in April 2020, but it too had fallen below 20% by June and July, so even that measure did not sit in the 20–30% range during the summer months. (cnbc.com)

However, Friedberg explicitly framed the 20–30% figure as conditional: it would occur “if we keep this up,” i.e., if the then‑current level of economic shutdown persisted. In practice, that condition did not hold. All U.S. states began phased reopenings in late April and May, and by June 2020 every state had at least partially lifted stay‑at‑home orders and business closures, substantially relaxing the economic shutdown compared with March. (pmc.ncbi.nlm.nih.gov) Because the severe nationwide lockdown did not continue through the summer, we never observe the counterfactual world Friedberg was talking about.

So, as a prediction about what actually happened, unemployment did not reach 20–30% by summer 2020; but as a strictly conditional “if the shutdown at March levels continues” forecast, the antecedent never occurred, making its truth value inherently indeterminate. For that reason, the fairest grading is ambiguous rather than clearly right or wrong.