The recovery from the devastating 2007-2009 financial crisis has been unusually lengthy. The latest growth stretch has already lasted 96 months, and if the poll predictions come true it would mark the longest economic expansion in more than 150 years.

Growth has still not picked up as quickly as thought recently, leading forecasters to lower expectations again slightly in the poll of more than 100 economists taken Aug. 7-10.

Still, the U.S. expansion has more than two years to go, according to 34 of 57 economists who answered an additional question on the business cycle. Of those economists, 21 said it would last two to three years and 13 said more than three years.

"Expansions don't go on forever," said Sam Bullard, senior economist at Wells Fargo, who said there was another two to three years to go. "Steady, moderate growth looks like it could stay in place for a while."

The remaining 23 respondents said the expansion would only last one to two years. None of the economists, based in the United States, Canada and Europe, expected it to end within a year.

U.S. President Donald Trump's administration aims to boost annual growth to 3 percent, mainly through sweeping tax cuts. But with the failure to repeal and replace the Affordable Care Act, significant fiscal stimulus appears less likely and the economy has shown no signs of accelerating to meet that target.

Predictions pointed to continued sluggish average growth in the current economic cycle compared with previous cycles of this length, based on National Bureau of Economic Research data. (www.nber.org/cycles.html)

GDP likely grew at a 2.6 percent annualized pace in the second quarter, down from 2.7 percent in the July poll. But the trend has yet to break away from roughly 2 percent.

The latest poll suggests 2.1 percent to 2.5 percent growth each quarter to the end of next year, slightly down from the 2.2 percent to 2.5 percent predicted the previous month. But growth has not been that steady during this expansion and generally is not in any economy.

The modest outlook was still broadly explained by slower spending due to sluggish wage growth even though the economy is close to full employment. Expectations for tax cuts from the Trump administration are also fading.

While that has not deterred U.S. stock markets, which have been setting record highs all year, it has pushed the dollar down nearly 9 percent against a basket of currencies.

Inflation forecasts have remained lukewarm, with the Federal Reserve's preferred gauge, the core PCE price index, not expected to reach the central bank's 2 percent target until the final quarter of 2018.

Core PCE inflation was forecast to average 1.5 percent to 1.6 percent each quarter from here until the end of 2017.

AUTUMN IT IS

Despite that subdued inflation outlook, the Fed is still expected to announce steps to start shrinking its more than $4 trillion balance sheet in September, according to 94 of 100 economists in the Reuters poll.

Five respondents said the announcement would be some time in the final three months of this year and one said early next year.

The poll also predicted the Fed would raise interest rates by 25 basis points in October or December, taking the fed funds rate to a range of 1.25 percent to 1.50 percent. The Fed is expected to follow up with three more rate hikes of the same amount in 2018.

"What the Fed is doing right now is saying the healthy economy combined with strong financial conditions more than make up for the disappointment in inflation," said Ethan Harris, head of global economics at Bank of America Merrill Lynch.

"They are very likely to announce their balance sheet shrinkage in September and see better-than-even odds they will even hike in December."

When asked if the Fed should start shrinking its balance sheet before inflation hits its target, 55 of 62 economists said "yes."

"Yes, they need to lower the balance sheet given emergency conditions are absent," said Wells Fargo's Bullard.

But not everyone agrees. Some economists worry the envisioned pace of Fed tightening could hurt the economy, especially given the anemic nature of the recovery.

"Clearly the idea here is to get ready for the end of the (economic) cycle," said Harris.

"In my view, hiking now so you can cut later is not the right way to think about it. You hike late so you can hike more. Because (if) you get inflation higher, then the hikes aren't that painful to the economy."