SAN FRANCISCO (Reuters) - An economic adviser at the Federal Reserve Bank of San Francisco known for his research into productivity trends published an analysis Monday that left open the possibility that a recent surge in productivity may not necessarily fade as it so often has in the past.

John Fernald, economist emeritus at the San Francisco Fed and a professor at INSEAD in Fountainebleau, France, has long been cautious about extrapolating from short-term trends to conclude that the U.S. is entering a new period of breakout productivity growth like that from 1995 to 2004, which enabled big economic gains without inflation.

Some Fed policymakers and analysts have expressed the hope that generative artificial intelligence and other innovations may already be setting the nation up for such a period, which lays the groundwork for improvements in living standards.

Most of Fernald's latest piece in the regional Fed's Economic Letter was consistent with a more skeptical view, detailing how a jump in per-worker output at the start of the COVID-19 pandemic quickly faded and longer-term slower trends reasserted themselves.

"This pandemic boom-and-bust in productivity growth was a predictable cyclical response overlaid on a broad continuation of the underlying slow growth pace," Fernald and co-authors Huiyu Li, Brigid Meisenbacher and Aren Yalcin wrote, noting that productivity growth surged and faded during the Great Recession as well.

Productivity growth is now running close to, although slightly higher, than has been typical since 2004, they found.

And yet, they concluded, "there are some reasons for optimism," including recent official data revisions that show faster productivity growth since the pandemic than had been previously estimated.

"Much is still uncertain about the productivity effects of emerging technologies like generative artificial intelligence, which will only be revealed over time, as the economy continues to evolve in the aftermath of the pandemic."